Upload fine-tuned model - BLEU: 0.2320
Browse files- .gitattributes +6 -0
- README.md +4 -48
- baseline_results.json +3 -3
- checkpoint-11000/config.json +54 -0
- checkpoint-11000/generation_config.json +17 -0
- checkpoint-11000/model.safetensors +3 -0
- checkpoint-11000/optimizer.pt +3 -0
- checkpoint-11000/rng_state.pth +3 -0
- checkpoint-11000/scaler.pt +3 -0
- checkpoint-11000/scheduler.pt +3 -0
- checkpoint-11000/source.spm +3 -0
- checkpoint-11000/special_tokens_map.json +5 -0
- checkpoint-11000/target.spm +3 -0
- checkpoint-11000/tokenizer_config.json +39 -0
- checkpoint-11000/trainer_state.json +1704 -0
- checkpoint-11000/training_args.bin +3 -0
- checkpoint-11000/vocab.json +0 -0
- checkpoint-11416/config.json +54 -0
- checkpoint-11416/generation_config.json +17 -0
- checkpoint-11416/model.safetensors +3 -0
- checkpoint-11416/optimizer.pt +3 -0
- checkpoint-11416/rng_state.pth +3 -0
- checkpoint-11416/scaler.pt +3 -0
- checkpoint-11416/scheduler.pt +3 -0
- checkpoint-11416/source.spm +3 -0
- checkpoint-11416/special_tokens_map.json +5 -0
- checkpoint-11416/target.spm +3 -0
- checkpoint-11416/tokenizer_config.json +39 -0
- checkpoint-11416/trainer_state.json +1760 -0
- checkpoint-11416/training_args.bin +3 -0
- checkpoint-11416/vocab.json +0 -0
- checkpoint-384/training_args.bin +1 -1
- emergency_checkpoint/config.json +54 -0
- emergency_checkpoint/generation_config.json +17 -0
- emergency_checkpoint/model.safetensors +3 -0
- emergency_checkpoint/source.spm +3 -0
- emergency_checkpoint/special_tokens_map.json +5 -0
- emergency_checkpoint/target.spm +3 -0
- emergency_checkpoint/tokenizer_config.json +39 -0
- emergency_checkpoint/vocab.json +0 -0
- test_results.json +3 -3
- train_results.json +3 -3
- training_args.bin +1 -1
- translation_config.json +3 -3
.gitattributes
CHANGED
|
@@ -51,3 +51,9 @@ checkpoint-144/source.spm filter=lfs diff=lfs merge=lfs -text
|
|
| 51 |
checkpoint-144/target.spm filter=lfs diff=lfs merge=lfs -text
|
| 52 |
checkpoint-384/source.spm filter=lfs diff=lfs merge=lfs -text
|
| 53 |
checkpoint-384/target.spm filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 51 |
checkpoint-144/target.spm filter=lfs diff=lfs merge=lfs -text
|
| 52 |
checkpoint-384/source.spm filter=lfs diff=lfs merge=lfs -text
|
| 53 |
checkpoint-384/target.spm filter=lfs diff=lfs merge=lfs -text
|
| 54 |
+
checkpoint-11000/source.spm filter=lfs diff=lfs merge=lfs -text
|
| 55 |
+
checkpoint-11000/target.spm filter=lfs diff=lfs merge=lfs -text
|
| 56 |
+
checkpoint-11416/source.spm filter=lfs diff=lfs merge=lfs -text
|
| 57 |
+
checkpoint-11416/target.spm filter=lfs diff=lfs merge=lfs -text
|
| 58 |
+
emergency_checkpoint/source.spm filter=lfs diff=lfs merge=lfs -text
|
| 59 |
+
emergency_checkpoint/target.spm filter=lfs diff=lfs merge=lfs -text
|
README.md
CHANGED
|
@@ -22,21 +22,13 @@ metrics:
|
|
| 22 |
|
| 23 |
## Model Description
|
| 24 |
|
| 25 |
-
This model is a fine-tuned version of `Helsinki-NLP/opus-mt-mul-en` for Swahili-to-English translation, specifically optimized for child helpline call transcriptions in East Africa
|
| 26 |
|
| 27 |
**Developed by:** BITZ IT Consulting Ltd
|
| 28 |
**Project:** OpenCHS (Open Child Helpline System)
|
| 29 |
**Funded by:** UNICEF Venture Fund
|
| 30 |
**License:** Apache 2.0
|
| 31 |
|
| 32 |
-
## Training Data
|
| 33 |
-
|
| 34 |
-
The model was fine-tuned on a combination of:
|
| 35 |
-
- NLLB Swahili-English parallel corpus (high quality, weight: 1.0)
|
| 36 |
-
- CCAligned web-crawled parallel data (supplementary, weight: 0.7)
|
| 37 |
-
- Total training samples: Approximately 50,000+ sentence pairs
|
| 38 |
-
- Domain focus: Conversational Swahili from helpline contexts
|
| 39 |
-
|
| 40 |
## Performance
|
| 41 |
|
| 42 |
### Test Set (General Translation)
|
|
@@ -49,34 +41,21 @@ The model was fine-tuned on a combination of:
|
|
| 49 |
- **Domain chrF:** 2.80
|
| 50 |
- **Domain COMET-QE:** 0.0000
|
| 51 |
|
| 52 |
-
*Domain metrics are evaluated on real 10-minute call transcriptions from child helplines.*
|
| 53 |
-
|
| 54 |
## Intended Use
|
| 55 |
|
| 56 |
-
**Primary Use Case:** Translating Swahili helpline call transcriptions to English for
|
| 57 |
-
- Case documentation and reporting
|
| 58 |
-
- Quality assurance and supervision
|
| 59 |
-
- Cross-border case referrals
|
| 60 |
-
- Data analysis and insights
|
| 61 |
|
| 62 |
**Languages:** Swahili (source) → English (target)
|
| 63 |
|
| 64 |
-
**Limitations:**
|
| 65 |
-
- Optimized for conversational Swahili (East African dialects)
|
| 66 |
-
- May not perform well on formal/literary Swahili
|
| 67 |
-
- Best for text lengths under 512 tokens
|
| 68 |
-
- Requires post-editing for critical use cases
|
| 69 |
-
|
| 70 |
## Usage
|
| 71 |
|
| 72 |
```python
|
| 73 |
from transformers import MarianTokenizer, MarianMTModel
|
| 74 |
|
| 75 |
-
model_name = "
|
| 76 |
tokenizer = MarianTokenizer.from_pretrained(model_name)
|
| 77 |
model = MarianMTModel.from_pretrained(model_name)
|
| 78 |
|
| 79 |
-
# Translate
|
| 80 |
swahili_text = "Habari za asubuhi. Ninaitwa Amina na nina miaka 14."
|
| 81 |
inputs = tokenizer(swahili_text, return_tensors="pt", padding=True)
|
| 82 |
outputs = model.generate(**inputs, num_beams=5, max_length=256)
|
|
@@ -88,33 +67,10 @@ print(translation)
|
|
| 88 |
|
| 89 |
**Base Model:** Helsinki-NLP/opus-mt-mul-en
|
| 90 |
**Training Epochs:** 8
|
| 91 |
-
**Batch Size:** 128
|
| 92 |
**Learning Rate:** 3e-05
|
| 93 |
-
**Optimizer:** AdamW
|
| 94 |
**Hardware:** NVIDIA GPU with FP16 mixed precision
|
| 95 |
|
| 96 |
-
## Evaluation Methodology
|
| 97 |
-
|
| 98 |
-
1. **Test Set:** Random 5% split from training distribution
|
| 99 |
-
2. **Domain Evaluation:** Held-out set of real helpline call transcriptions (10-min calls)
|
| 100 |
-
3. **Metrics:** BLEU, chrF, COMET-QE (quality estimation)
|
| 101 |
-
|
| 102 |
-
## Citation
|
| 103 |
-
|
| 104 |
-
```bibtex
|
| 105 |
-
@software{openchs_translation_2025,
|
| 106 |
-
author = {BITZ IT Consulting Ltd},
|
| 107 |
-
title = {Swahili-English Translation Model for OpenCHS},
|
| 108 |
-
year = {2025},
|
| 109 |
-
publisher = {HuggingFace},
|
| 110 |
-
url = {https://huggingface.co/YOUR_USERNAME/brendaogutu/sw-en-translation-v1}
|
| 111 |
-
}
|
| 112 |
-
```
|
| 113 |
-
|
| 114 |
-
## Contact
|
| 115 |
-
|
| 116 |
-
For questions or issues, please contact: brenda@openchs.org
|
| 117 |
-
|
| 118 |
---
|
| 119 |
|
| 120 |
*This model is part of the OpenCHS project supporting child helpline services across East Africa.*
|
|
|
|
| 22 |
|
| 23 |
## Model Description
|
| 24 |
|
| 25 |
+
This model is a fine-tuned version of `Helsinki-NLP/opus-mt-mul-en` for Swahili-to-English translation, specifically optimized for child helpline call transcriptions in East Africa.
|
| 26 |
|
| 27 |
**Developed by:** BITZ IT Consulting Ltd
|
| 28 |
**Project:** OpenCHS (Open Child Helpline System)
|
| 29 |
**Funded by:** UNICEF Venture Fund
|
| 30 |
**License:** Apache 2.0
|
| 31 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 32 |
## Performance
|
| 33 |
|
| 34 |
### Test Set (General Translation)
|
|
|
|
| 41 |
- **Domain chrF:** 2.80
|
| 42 |
- **Domain COMET-QE:** 0.0000
|
| 43 |
|
|
|
|
|
|
|
| 44 |
## Intended Use
|
| 45 |
|
| 46 |
+
**Primary Use Case:** Translating Swahili helpline call transcriptions to English for case documentation, quality assurance, and cross-border referrals.
|
|
|
|
|
|
|
|
|
|
|
|
|
| 47 |
|
| 48 |
**Languages:** Swahili (source) → English (target)
|
| 49 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 50 |
## Usage
|
| 51 |
|
| 52 |
```python
|
| 53 |
from transformers import MarianTokenizer, MarianMTModel
|
| 54 |
|
| 55 |
+
model_name = "brendaogutu/sw-en-translation-v1"
|
| 56 |
tokenizer = MarianTokenizer.from_pretrained(model_name)
|
| 57 |
model = MarianMTModel.from_pretrained(model_name)
|
| 58 |
|
|
|
|
| 59 |
swahili_text = "Habari za asubuhi. Ninaitwa Amina na nina miaka 14."
|
| 60 |
inputs = tokenizer(swahili_text, return_tensors="pt", padding=True)
|
| 61 |
outputs = model.generate(**inputs, num_beams=5, max_length=256)
|
|
|
|
| 67 |
|
| 68 |
**Base Model:** Helsinki-NLP/opus-mt-mul-en
|
| 69 |
**Training Epochs:** 8
|
| 70 |
+
**Batch Size:** 128
|
| 71 |
**Learning Rate:** 3e-05
|
|
|
|
| 72 |
**Hardware:** NVIDIA GPU with FP16 mixed precision
|
| 73 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 74 |
---
|
| 75 |
|
| 76 |
*This model is part of the OpenCHS project supporting child helpline services across East Africa.*
|
baseline_results.json
CHANGED
|
@@ -3,9 +3,9 @@
|
|
| 3 |
"eval_model_preparation_time": 0.0023,
|
| 4 |
"eval_bleu": 0.03171354860381341,
|
| 5 |
"eval_chrf": 20.030925699576805,
|
| 6 |
-
"eval_runtime":
|
| 7 |
-
"eval_samples_per_second":
|
| 8 |
-
"eval_steps_per_second": 0.
|
| 9 |
"eval_comet": 0.0341746505332392,
|
| 10 |
"eval_comet_std": 0.1127765594779826,
|
| 11 |
"baseline_domain_bleu": 0.0,
|
|
|
|
| 3 |
"eval_model_preparation_time": 0.0023,
|
| 4 |
"eval_bleu": 0.03171354860381341,
|
| 5 |
"eval_chrf": 20.030925699576805,
|
| 6 |
+
"eval_runtime": 10.9008,
|
| 7 |
+
"eval_samples_per_second": 33.025,
|
| 8 |
+
"eval_steps_per_second": 0.275,
|
| 9 |
"eval_comet": 0.0341746505332392,
|
| 10 |
"eval_comet_std": 0.1127765594779826,
|
| 11 |
"baseline_domain_bleu": 0.0,
|
checkpoint-11000/config.json
ADDED
|
@@ -0,0 +1,54 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"activation_dropout": 0.0,
|
| 3 |
+
"activation_function": "swish",
|
| 4 |
+
"add_bias_logits": false,
|
| 5 |
+
"add_final_layer_norm": false,
|
| 6 |
+
"architectures": [
|
| 7 |
+
"MarianMTModel"
|
| 8 |
+
],
|
| 9 |
+
"attention_dropout": 0.0,
|
| 10 |
+
"classif_dropout": 0.0,
|
| 11 |
+
"classifier_dropout": 0.0,
|
| 12 |
+
"d_model": 512,
|
| 13 |
+
"decoder_attention_heads": 8,
|
| 14 |
+
"decoder_ffn_dim": 2048,
|
| 15 |
+
"decoder_layerdrop": 0.0,
|
| 16 |
+
"decoder_layers": 6,
|
| 17 |
+
"decoder_start_token_id": 64171,
|
| 18 |
+
"decoder_vocab_size": 64172,
|
| 19 |
+
"dropout": 0.1,
|
| 20 |
+
"dtype": "float32",
|
| 21 |
+
"encoder_attention_heads": 8,
|
| 22 |
+
"encoder_ffn_dim": 2048,
|
| 23 |
+
"encoder_layerdrop": 0.0,
|
| 24 |
+
"encoder_layers": 6,
|
| 25 |
+
"eos_token_id": 0,
|
| 26 |
+
"extra_pos_embeddings": 64172,
|
| 27 |
+
"forced_eos_token_id": 0,
|
| 28 |
+
"id2label": {
|
| 29 |
+
"0": "LABEL_0",
|
| 30 |
+
"1": "LABEL_1",
|
| 31 |
+
"2": "LABEL_2"
|
| 32 |
+
},
|
| 33 |
+
"init_std": 0.02,
|
| 34 |
+
"is_encoder_decoder": true,
|
| 35 |
+
"label2id": {
|
| 36 |
+
"LABEL_0": 0,
|
| 37 |
+
"LABEL_1": 1,
|
| 38 |
+
"LABEL_2": 2
|
| 39 |
+
},
|
| 40 |
+
"max_length": null,
|
| 41 |
+
"max_position_embeddings": 512,
|
| 42 |
+
"model_type": "marian",
|
| 43 |
+
"normalize_before": false,
|
| 44 |
+
"normalize_embedding": false,
|
| 45 |
+
"num_beams": null,
|
| 46 |
+
"num_hidden_layers": 6,
|
| 47 |
+
"pad_token_id": 64171,
|
| 48 |
+
"scale_embedding": true,
|
| 49 |
+
"share_encoder_decoder_embeddings": true,
|
| 50 |
+
"static_position_embeddings": true,
|
| 51 |
+
"transformers_version": "4.57.1",
|
| 52 |
+
"use_cache": true,
|
| 53 |
+
"vocab_size": 64172
|
| 54 |
+
}
|
checkpoint-11000/generation_config.json
ADDED
|
@@ -0,0 +1,17 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bad_words_ids": [
|
| 3 |
+
[
|
| 4 |
+
64171
|
| 5 |
+
]
|
| 6 |
+
],
|
| 7 |
+
"decoder_start_token_id": 64171,
|
| 8 |
+
"eos_token_id": [
|
| 9 |
+
0
|
| 10 |
+
],
|
| 11 |
+
"forced_eos_token_id": 0,
|
| 12 |
+
"max_length": 512,
|
| 13 |
+
"num_beams": 6,
|
| 14 |
+
"pad_token_id": 64171,
|
| 15 |
+
"renormalize_logits": true,
|
| 16 |
+
"transformers_version": "4.57.1"
|
| 17 |
+
}
|
checkpoint-11000/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:968e4e015c98e23af3da62d2661b4ed129a2b3a040b7e9d159ec6b6e5b3b4def
|
| 3 |
+
size 308263984
|
checkpoint-11000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:09e02b7f67d4035c6b72de1fd4e1b7021e4a02bf70105fbcae92a9f3689f2f86
|
| 3 |
+
size 616171979
|
checkpoint-11000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:03d5c7f715d3b7f10519fd168eda0c46abe51325cc13804182c3d237388037cb
|
| 3 |
+
size 14645
|
checkpoint-11000/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:68ac27aa7b9ce48c2cbfca3c70fd196671a369fa834d26faaa9383ca02d093ea
|
| 3 |
+
size 1383
|
checkpoint-11000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d9e50bed4acf1a795f883e1907ad05d110427a168a769172a889bb22ac242c8a
|
| 3 |
+
size 1465
|
checkpoint-11000/source.spm
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c4a99ea3602b29fbf901ade8b93a45efa3d7c64eab8fc5fa812383efa327a87d
|
| 3 |
+
size 706917
|
checkpoint-11000/special_tokens_map.json
ADDED
|
@@ -0,0 +1,5 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"eos_token": "</s>",
|
| 3 |
+
"pad_token": "<pad>",
|
| 4 |
+
"unk_token": "<unk>"
|
| 5 |
+
}
|
checkpoint-11000/target.spm
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c6dce5fa58fcd7dde9e81e279b8c075bf42ee558278f73d6fb48e342029d7f19
|
| 3 |
+
size 791194
|
checkpoint-11000/tokenizer_config.json
ADDED
|
@@ -0,0 +1,39 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "</s>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<unk>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"64171": {
|
| 20 |
+
"content": "<pad>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
}
|
| 27 |
+
},
|
| 28 |
+
"clean_up_tokenization_spaces": false,
|
| 29 |
+
"eos_token": "</s>",
|
| 30 |
+
"extra_special_tokens": {},
|
| 31 |
+
"model_max_length": 512,
|
| 32 |
+
"pad_token": "<pad>",
|
| 33 |
+
"separate_vocabs": false,
|
| 34 |
+
"source_lang": "mul",
|
| 35 |
+
"sp_model_kwargs": {},
|
| 36 |
+
"target_lang": "eng",
|
| 37 |
+
"tokenizer_class": "MarianTokenizer",
|
| 38 |
+
"unk_token": "<unk>"
|
| 39 |
+
}
|
checkpoint-11000/trainer_state.json
ADDED
|
@@ -0,0 +1,1704 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 11000,
|
| 3 |
+
"best_metric": 0.4164085587117329,
|
| 4 |
+
"best_model_checkpoint": "models/finetuned-sw-en/checkpoint-11000",
|
| 5 |
+
"epoch": 7.708479327259986,
|
| 6 |
+
"eval_steps": 1000,
|
| 7 |
+
"global_step": 11000,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.0350385423966363,
|
| 14 |
+
"grad_norm": 140.77987670898438,
|
| 15 |
+
"learning_rate": 1.2872154115586689e-06,
|
| 16 |
+
"loss": 15.3091,
|
| 17 |
+
"step": 50
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.0700770847932726,
|
| 21 |
+
"grad_norm": 20.989227294921875,
|
| 22 |
+
"learning_rate": 2.6007005253940457e-06,
|
| 23 |
+
"loss": 7.7676,
|
| 24 |
+
"step": 100
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.1051156271899089,
|
| 28 |
+
"grad_norm": 5.4014387130737305,
|
| 29 |
+
"learning_rate": 3.914185639229422e-06,
|
| 30 |
+
"loss": 4.6742,
|
| 31 |
+
"step": 150
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.1401541695865452,
|
| 35 |
+
"grad_norm": 5.136999607086182,
|
| 36 |
+
"learning_rate": 5.227670753064799e-06,
|
| 37 |
+
"loss": 4.2477,
|
| 38 |
+
"step": 200
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.1751927119831815,
|
| 42 |
+
"grad_norm": 5.013916015625,
|
| 43 |
+
"learning_rate": 6.541155866900175e-06,
|
| 44 |
+
"loss": 4.0228,
|
| 45 |
+
"step": 250
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.2102312543798178,
|
| 49 |
+
"grad_norm": 4.574016571044922,
|
| 50 |
+
"learning_rate": 7.854640980735552e-06,
|
| 51 |
+
"loss": 3.8679,
|
| 52 |
+
"step": 300
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.2452697967764541,
|
| 56 |
+
"grad_norm": 4.371525287628174,
|
| 57 |
+
"learning_rate": 9.168126094570928e-06,
|
| 58 |
+
"loss": 3.7254,
|
| 59 |
+
"step": 350
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.2803083391730904,
|
| 63 |
+
"grad_norm": 4.383193492889404,
|
| 64 |
+
"learning_rate": 1.0481611208406305e-05,
|
| 65 |
+
"loss": 3.6216,
|
| 66 |
+
"step": 400
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.3153468815697267,
|
| 70 |
+
"grad_norm": 4.636805534362793,
|
| 71 |
+
"learning_rate": 1.1795096322241681e-05,
|
| 72 |
+
"loss": 3.5232,
|
| 73 |
+
"step": 450
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.350385423966363,
|
| 77 |
+
"grad_norm": 4.319335460662842,
|
| 78 |
+
"learning_rate": 1.3108581436077059e-05,
|
| 79 |
+
"loss": 3.4235,
|
| 80 |
+
"step": 500
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.3854239663629993,
|
| 84 |
+
"grad_norm": 3.9548702239990234,
|
| 85 |
+
"learning_rate": 1.4422066549912435e-05,
|
| 86 |
+
"loss": 3.3257,
|
| 87 |
+
"step": 550
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.4204625087596356,
|
| 91 |
+
"grad_norm": 3.984466314315796,
|
| 92 |
+
"learning_rate": 1.5735551663747813e-05,
|
| 93 |
+
"loss": 3.2771,
|
| 94 |
+
"step": 600
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.4555010511562719,
|
| 98 |
+
"grad_norm": 3.979240655899048,
|
| 99 |
+
"learning_rate": 1.7049036777583185e-05,
|
| 100 |
+
"loss": 3.2041,
|
| 101 |
+
"step": 650
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.4905395935529082,
|
| 105 |
+
"grad_norm": 3.8334100246429443,
|
| 106 |
+
"learning_rate": 1.8362521891418565e-05,
|
| 107 |
+
"loss": 3.1511,
|
| 108 |
+
"step": 700
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.5255781359495445,
|
| 112 |
+
"grad_norm": 3.752436637878418,
|
| 113 |
+
"learning_rate": 1.967600700525394e-05,
|
| 114 |
+
"loss": 3.1209,
|
| 115 |
+
"step": 750
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.5606166783461808,
|
| 119 |
+
"grad_norm": 3.773709535598755,
|
| 120 |
+
"learning_rate": 2.0989492119089317e-05,
|
| 121 |
+
"loss": 3.049,
|
| 122 |
+
"step": 800
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.5956552207428171,
|
| 126 |
+
"grad_norm": 3.9539151191711426,
|
| 127 |
+
"learning_rate": 2.2302977232924692e-05,
|
| 128 |
+
"loss": 3.0337,
|
| 129 |
+
"step": 850
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.6306937631394534,
|
| 133 |
+
"grad_norm": 3.7437591552734375,
|
| 134 |
+
"learning_rate": 2.3616462346760072e-05,
|
| 135 |
+
"loss": 2.9983,
|
| 136 |
+
"step": 900
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.6657323055360898,
|
| 140 |
+
"grad_norm": 3.707627773284912,
|
| 141 |
+
"learning_rate": 2.4929947460595444e-05,
|
| 142 |
+
"loss": 2.9624,
|
| 143 |
+
"step": 950
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.700770847932726,
|
| 147 |
+
"grad_norm": 3.768141508102417,
|
| 148 |
+
"learning_rate": 2.6243432574430824e-05,
|
| 149 |
+
"loss": 2.9167,
|
| 150 |
+
"step": 1000
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.700770847932726,
|
| 154 |
+
"eval_bleu": 0.3189127184480538,
|
| 155 |
+
"eval_chrf": 51.264833545919984,
|
| 156 |
+
"eval_loss": 2.8209688663482666,
|
| 157 |
+
"eval_model_preparation_time": 0.0023,
|
| 158 |
+
"eval_runtime": 221.3909,
|
| 159 |
+
"eval_samples_per_second": 97.059,
|
| 160 |
+
"eval_steps_per_second": 0.759,
|
| 161 |
+
"step": 1000
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 0.7358093903293623,
|
| 165 |
+
"grad_norm": 4.0691142082214355,
|
| 166 |
+
"learning_rate": 2.75569176882662e-05,
|
| 167 |
+
"loss": 2.9229,
|
| 168 |
+
"step": 1050
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 0.7708479327259986,
|
| 172 |
+
"grad_norm": 3.470956325531006,
|
| 173 |
+
"learning_rate": 2.8870402802101576e-05,
|
| 174 |
+
"loss": 2.8988,
|
| 175 |
+
"step": 1100
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 0.8058864751226349,
|
| 179 |
+
"grad_norm": 3.2942214012145996,
|
| 180 |
+
"learning_rate": 2.9999965638050107e-05,
|
| 181 |
+
"loss": 2.9036,
|
| 182 |
+
"step": 1150
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.8409250175192712,
|
| 186 |
+
"grad_norm": 3.5197594165802,
|
| 187 |
+
"learning_rate": 2.999772164915165e-05,
|
| 188 |
+
"loss": 2.8677,
|
| 189 |
+
"step": 1200
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.8759635599159075,
|
| 193 |
+
"grad_norm": 3.3883650302886963,
|
| 194 |
+
"learning_rate": 2.999197193836303e-05,
|
| 195 |
+
"loss": 2.8463,
|
| 196 |
+
"step": 1250
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 0.9110021023125437,
|
| 200 |
+
"grad_norm": 3.5236427783966064,
|
| 201 |
+
"learning_rate": 2.998271784968085e-05,
|
| 202 |
+
"loss": 2.8359,
|
| 203 |
+
"step": 1300
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 0.9460406447091801,
|
| 207 |
+
"grad_norm": 3.7130653858184814,
|
| 208 |
+
"learning_rate": 2.9969961546251093e-05,
|
| 209 |
+
"loss": 2.8076,
|
| 210 |
+
"step": 1350
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 0.9810791871058164,
|
| 214 |
+
"grad_norm": 3.4754154682159424,
|
| 215 |
+
"learning_rate": 2.9953706009863472e-05,
|
| 216 |
+
"loss": 2.8054,
|
| 217 |
+
"step": 1400
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 1.0161177295024526,
|
| 221 |
+
"grad_norm": 3.351656675338745,
|
| 222 |
+
"learning_rate": 2.9933955040254443e-05,
|
| 223 |
+
"loss": 2.7778,
|
| 224 |
+
"step": 1450
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 1.051156271899089,
|
| 228 |
+
"grad_norm": 3.2852399349212646,
|
| 229 |
+
"learning_rate": 2.9910713254219016e-05,
|
| 230 |
+
"loss": 2.7308,
|
| 231 |
+
"step": 1500
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 1.0861948142957254,
|
| 235 |
+
"grad_norm": 3.4650189876556396,
|
| 236 |
+
"learning_rate": 2.9883986084531597e-05,
|
| 237 |
+
"loss": 2.7357,
|
| 238 |
+
"step": 1550
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 1.1212333566923616,
|
| 242 |
+
"grad_norm": 3.4110779762268066,
|
| 243 |
+
"learning_rate": 2.985377977867602e-05,
|
| 244 |
+
"loss": 2.7242,
|
| 245 |
+
"step": 1600
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 1.156271899088998,
|
| 249 |
+
"grad_norm": 3.506265163421631,
|
| 250 |
+
"learning_rate": 2.982010139738526e-05,
|
| 251 |
+
"loss": 2.707,
|
| 252 |
+
"step": 1650
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 1.1913104414856341,
|
| 256 |
+
"grad_norm": 3.2563109397888184,
|
| 257 |
+
"learning_rate": 2.978295881299094e-05,
|
| 258 |
+
"loss": 2.723,
|
| 259 |
+
"step": 1700
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"epoch": 1.2263489838822705,
|
| 263 |
+
"grad_norm": 3.262761354446411,
|
| 264 |
+
"learning_rate": 2.9742360707583205e-05,
|
| 265 |
+
"loss": 2.7156,
|
| 266 |
+
"step": 1750
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 1.2613875262789067,
|
| 270 |
+
"grad_norm": 3.3031880855560303,
|
| 271 |
+
"learning_rate": 2.969831657098125e-05,
|
| 272 |
+
"loss": 2.7032,
|
| 273 |
+
"step": 1800
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 1.296426068675543,
|
| 277 |
+
"grad_norm": 3.6921067237854004,
|
| 278 |
+
"learning_rate": 2.9650836698515077e-05,
|
| 279 |
+
"loss": 2.7066,
|
| 280 |
+
"step": 1850
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 1.3314646110721795,
|
| 284 |
+
"grad_norm": 3.372687578201294,
|
| 285 |
+
"learning_rate": 2.9599932188618993e-05,
|
| 286 |
+
"loss": 2.6926,
|
| 287 |
+
"step": 1900
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 1.3665031534688157,
|
| 291 |
+
"grad_norm": 3.5358870029449463,
|
| 292 |
+
"learning_rate": 2.9545614940237294e-05,
|
| 293 |
+
"loss": 2.6876,
|
| 294 |
+
"step": 1950
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 1.401541695865452,
|
| 298 |
+
"grad_norm": 3.1628262996673584,
|
| 299 |
+
"learning_rate": 2.9487897650042928e-05,
|
| 300 |
+
"loss": 2.6921,
|
| 301 |
+
"step": 2000
|
| 302 |
+
},
|
| 303 |
+
{
|
| 304 |
+
"epoch": 1.401541695865452,
|
| 305 |
+
"eval_bleu": 0.3710693297073462,
|
| 306 |
+
"eval_chrf": 56.162355444042014,
|
| 307 |
+
"eval_loss": 2.6536850929260254,
|
| 308 |
+
"eval_model_preparation_time": 0.0023,
|
| 309 |
+
"eval_runtime": 226.613,
|
| 310 |
+
"eval_samples_per_second": 94.822,
|
| 311 |
+
"eval_steps_per_second": 0.741,
|
| 312 |
+
"step": 2000
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 1.4365802382620882,
|
| 316 |
+
"grad_norm": 3.2792763710021973,
|
| 317 |
+
"learning_rate": 2.942679380946962e-05,
|
| 318 |
+
"loss": 2.6852,
|
| 319 |
+
"step": 2050
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 1.4716187806587246,
|
| 323 |
+
"grad_norm": 3.4099667072296143,
|
| 324 |
+
"learning_rate": 2.9362317701558252e-05,
|
| 325 |
+
"loss": 2.672,
|
| 326 |
+
"step": 2100
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 1.5066573230553608,
|
| 330 |
+
"grad_norm": 3.4033470153808594,
|
| 331 |
+
"learning_rate": 2.9294484397618206e-05,
|
| 332 |
+
"loss": 2.6847,
|
| 333 |
+
"step": 2150
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 1.5416958654519972,
|
| 337 |
+
"grad_norm": 3.6894521713256836,
|
| 338 |
+
"learning_rate": 2.9223309753704418e-05,
|
| 339 |
+
"loss": 2.6789,
|
| 340 |
+
"step": 2200
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 1.5767344078486336,
|
| 344 |
+
"grad_norm": 3.1698200702667236,
|
| 345 |
+
"learning_rate": 2.9148810406911045e-05,
|
| 346 |
+
"loss": 2.6632,
|
| 347 |
+
"step": 2250
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 1.6117729502452698,
|
| 351 |
+
"grad_norm": 3.2145323753356934,
|
| 352 |
+
"learning_rate": 2.9071003771482533e-05,
|
| 353 |
+
"loss": 2.68,
|
| 354 |
+
"step": 2300
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 1.646811492641906,
|
| 358 |
+
"grad_norm": 3.1463518142700195,
|
| 359 |
+
"learning_rate": 2.898990803474302e-05,
|
| 360 |
+
"loss": 2.6609,
|
| 361 |
+
"step": 2350
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 1.6818500350385424,
|
| 365 |
+
"grad_norm": 3.1688807010650635,
|
| 366 |
+
"learning_rate": 2.8905542152845054e-05,
|
| 367 |
+
"loss": 2.6562,
|
| 368 |
+
"step": 2400
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 1.7168885774351788,
|
| 372 |
+
"grad_norm": 3.3895840644836426,
|
| 373 |
+
"learning_rate": 2.881792584633859e-05,
|
| 374 |
+
"loss": 2.6593,
|
| 375 |
+
"step": 2450
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 1.751927119831815,
|
| 379 |
+
"grad_norm": 3.324277639389038,
|
| 380 |
+
"learning_rate": 2.87270795955613e-05,
|
| 381 |
+
"loss": 2.6303,
|
| 382 |
+
"step": 2500
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 1.7869656622284513,
|
| 386 |
+
"grad_norm": 3.1676182746887207,
|
| 387 |
+
"learning_rate": 2.8633024635851283e-05,
|
| 388 |
+
"loss": 2.6473,
|
| 389 |
+
"step": 2550
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 1.8220042046250877,
|
| 393 |
+
"grad_norm": 3.058241128921509,
|
| 394 |
+
"learning_rate": 2.8535782952583308e-05,
|
| 395 |
+
"loss": 2.647,
|
| 396 |
+
"step": 2600
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 1.857042747021724,
|
| 400 |
+
"grad_norm": 3.310572624206543,
|
| 401 |
+
"learning_rate": 2.8435377276029724e-05,
|
| 402 |
+
"loss": 2.6562,
|
| 403 |
+
"step": 2650
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 1.89208128941836,
|
| 407 |
+
"grad_norm": 3.1375229358673096,
|
| 408 |
+
"learning_rate": 2.8331831076047245e-05,
|
| 409 |
+
"loss": 2.6474,
|
| 410 |
+
"step": 2700
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 1.9271198318149965,
|
| 414 |
+
"grad_norm": 3.0486273765563965,
|
| 415 |
+
"learning_rate": 2.822516855659088e-05,
|
| 416 |
+
"loss": 2.6356,
|
| 417 |
+
"step": 2750
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 1.9621583742116329,
|
| 421 |
+
"grad_norm": 3.2743468284606934,
|
| 422 |
+
"learning_rate": 2.8115414650056225e-05,
|
| 423 |
+
"loss": 2.639,
|
| 424 |
+
"step": 2800
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 1.997196916608269,
|
| 428 |
+
"grad_norm": 3.156439781188965,
|
| 429 |
+
"learning_rate": 2.8002595011451535e-05,
|
| 430 |
+
"loss": 2.6246,
|
| 431 |
+
"step": 2850
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 2.032235459004905,
|
| 435 |
+
"grad_norm": 3.180952548980713,
|
| 436 |
+
"learning_rate": 2.7886736012400844e-05,
|
| 437 |
+
"loss": 2.5383,
|
| 438 |
+
"step": 2900
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 2.067274001401542,
|
| 442 |
+
"grad_norm": 3.3652632236480713,
|
| 443 |
+
"learning_rate": 2.7767864734979593e-05,
|
| 444 |
+
"loss": 2.5578,
|
| 445 |
+
"step": 2950
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 2.102312543798178,
|
| 449 |
+
"grad_norm": 3.169829845428467,
|
| 450 |
+
"learning_rate": 2.7646008965384187e-05,
|
| 451 |
+
"loss": 2.5472,
|
| 452 |
+
"step": 3000
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 2.102312543798178,
|
| 456 |
+
"eval_bleu": 0.3885753538883607,
|
| 457 |
+
"eval_chrf": 57.86289492703291,
|
| 458 |
+
"eval_loss": 2.604308605194092,
|
| 459 |
+
"eval_model_preparation_time": 0.0023,
|
| 460 |
+
"eval_runtime": 251.154,
|
| 461 |
+
"eval_samples_per_second": 85.557,
|
| 462 |
+
"eval_steps_per_second": 0.669,
|
| 463 |
+
"step": 3000
|
| 464 |
+
},
|
| 465 |
+
{
|
| 466 |
+
"epoch": 2.137351086194814,
|
| 467 |
+
"grad_norm": 3.187664747238159,
|
| 468 |
+
"learning_rate": 2.7521197187436958e-05,
|
| 469 |
+
"loss": 2.5555,
|
| 470 |
+
"step": 3050
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 2.172389628591451,
|
| 474 |
+
"grad_norm": 3.2646875381469727,
|
| 475 |
+
"learning_rate": 2.7393458575928085e-05,
|
| 476 |
+
"loss": 2.5691,
|
| 477 |
+
"step": 3100
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 2.207428170988087,
|
| 481 |
+
"grad_norm": 3.1656856536865234,
|
| 482 |
+
"learning_rate": 2.7262822989795935e-05,
|
| 483 |
+
"loss": 2.5514,
|
| 484 |
+
"step": 3150
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 2.242466713384723,
|
| 488 |
+
"grad_norm": 2.9856343269348145,
|
| 489 |
+
"learning_rate": 2.712932096514758e-05,
|
| 490 |
+
"loss": 2.5438,
|
| 491 |
+
"step": 3200
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 2.2775052557813593,
|
| 495 |
+
"grad_norm": 3.210883378982544,
|
| 496 |
+
"learning_rate": 2.699298370812093e-05,
|
| 497 |
+
"loss": 2.5334,
|
| 498 |
+
"step": 3250
|
| 499 |
+
},
|
| 500 |
+
{
|
| 501 |
+
"epoch": 2.312543798177996,
|
| 502 |
+
"grad_norm": 3.069847345352173,
|
| 503 |
+
"learning_rate": 2.685384308759031e-05,
|
| 504 |
+
"loss": 2.5514,
|
| 505 |
+
"step": 3300
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 2.347582340574632,
|
| 509 |
+
"grad_norm": 3.252610445022583,
|
| 510 |
+
"learning_rate": 2.6711931627717095e-05,
|
| 511 |
+
"loss": 2.5626,
|
| 512 |
+
"step": 3350
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 2.3826208829712683,
|
| 516 |
+
"grad_norm": 3.202777862548828,
|
| 517 |
+
"learning_rate": 2.6567282500347186e-05,
|
| 518 |
+
"loss": 2.5462,
|
| 519 |
+
"step": 3400
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 2.417659425367905,
|
| 523 |
+
"grad_norm": 3.1640210151672363,
|
| 524 |
+
"learning_rate": 2.641992951725707e-05,
|
| 525 |
+
"loss": 2.5506,
|
| 526 |
+
"step": 3450
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 2.452697967764541,
|
| 530 |
+
"grad_norm": 3.1900179386138916,
|
| 531 |
+
"learning_rate": 2.6269907122250307e-05,
|
| 532 |
+
"loss": 2.555,
|
| 533 |
+
"step": 3500
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 2.4877365101611773,
|
| 537 |
+
"grad_norm": 2.921445846557617,
|
| 538 |
+
"learning_rate": 2.611725038310629e-05,
|
| 539 |
+
"loss": 2.5547,
|
| 540 |
+
"step": 3550
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 2.5227750525578134,
|
| 544 |
+
"grad_norm": 2.9808526039123535,
|
| 545 |
+
"learning_rate": 2.5961994983383144e-05,
|
| 546 |
+
"loss": 2.5642,
|
| 547 |
+
"step": 3600
|
| 548 |
+
},
|
| 549 |
+
{
|
| 550 |
+
"epoch": 2.55781359495445,
|
| 551 |
+
"grad_norm": 3.357646942138672,
|
| 552 |
+
"learning_rate": 2.5804177214076664e-05,
|
| 553 |
+
"loss": 2.5317,
|
| 554 |
+
"step": 3650
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 2.592852137351086,
|
| 558 |
+
"grad_norm": 3.1746153831481934,
|
| 559 |
+
"learning_rate": 2.5643833965137296e-05,
|
| 560 |
+
"loss": 2.5631,
|
| 561 |
+
"step": 3700
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 2.6278906797477224,
|
| 565 |
+
"grad_norm": 3.3090434074401855,
|
| 566 |
+
"learning_rate": 2.548100271684711e-05,
|
| 567 |
+
"loss": 2.5374,
|
| 568 |
+
"step": 3750
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 2.662929222144359,
|
| 572 |
+
"grad_norm": 3.5559844970703125,
|
| 573 |
+
"learning_rate": 2.531572153105876e-05,
|
| 574 |
+
"loss": 2.5469,
|
| 575 |
+
"step": 3800
|
| 576 |
+
},
|
| 577 |
+
{
|
| 578 |
+
"epoch": 2.697967764540995,
|
| 579 |
+
"grad_norm": 3.269394636154175,
|
| 580 |
+
"learning_rate": 2.5148029042298514e-05,
|
| 581 |
+
"loss": 2.5343,
|
| 582 |
+
"step": 3850
|
| 583 |
+
},
|
| 584 |
+
{
|
| 585 |
+
"epoch": 2.7330063069376314,
|
| 586 |
+
"grad_norm": 2.9711074829101562,
|
| 587 |
+
"learning_rate": 2.497796444873543e-05,
|
| 588 |
+
"loss": 2.5388,
|
| 589 |
+
"step": 3900
|
| 590 |
+
},
|
| 591 |
+
{
|
| 592 |
+
"epoch": 2.7680448493342675,
|
| 593 |
+
"grad_norm": 3.074882984161377,
|
| 594 |
+
"learning_rate": 2.480556750301877e-05,
|
| 595 |
+
"loss": 2.5361,
|
| 596 |
+
"step": 3950
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 2.803083391730904,
|
| 600 |
+
"grad_norm": 3.130532741546631,
|
| 601 |
+
"learning_rate": 2.463087850298579e-05,
|
| 602 |
+
"loss": 2.532,
|
| 603 |
+
"step": 4000
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 2.803083391730904,
|
| 607 |
+
"eval_bleu": 0.40072636523788635,
|
| 608 |
+
"eval_chrf": 58.74104751673321,
|
| 609 |
+
"eval_loss": 2.5759472846984863,
|
| 610 |
+
"eval_model_preparation_time": 0.0023,
|
| 611 |
+
"eval_runtime": 225.8985,
|
| 612 |
+
"eval_samples_per_second": 95.122,
|
| 613 |
+
"eval_steps_per_second": 0.744,
|
| 614 |
+
"step": 4000
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 2.8381219341275403,
|
| 618 |
+
"grad_norm": 3.0843958854675293,
|
| 619 |
+
"learning_rate": 2.4453938282242146e-05,
|
| 620 |
+
"loss": 2.547,
|
| 621 |
+
"step": 4050
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 2.8731604765241765,
|
| 625 |
+
"grad_norm": 2.9439711570739746,
|
| 626 |
+
"learning_rate": 2.4274788200616986e-05,
|
| 627 |
+
"loss": 2.5571,
|
| 628 |
+
"step": 4100
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 2.908199018920813,
|
| 632 |
+
"grad_norm": 3.01135516166687,
|
| 633 |
+
"learning_rate": 2.4093470134495104e-05,
|
| 634 |
+
"loss": 2.5365,
|
| 635 |
+
"step": 4150
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"epoch": 2.9432375613174493,
|
| 639 |
+
"grad_norm": 3.141979455947876,
|
| 640 |
+
"learning_rate": 2.3910026467028306e-05,
|
| 641 |
+
"loss": 2.5337,
|
| 642 |
+
"step": 4200
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 2.9782761037140855,
|
| 646 |
+
"grad_norm": 3.1309609413146973,
|
| 647 |
+
"learning_rate": 2.372450007822836e-05,
|
| 648 |
+
"loss": 2.5455,
|
| 649 |
+
"step": 4250
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 3.0133146461107216,
|
| 653 |
+
"grad_norm": 2.8289108276367188,
|
| 654 |
+
"learning_rate": 2.3536934334943747e-05,
|
| 655 |
+
"loss": 2.5126,
|
| 656 |
+
"step": 4300
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 3.0483531885073583,
|
| 660 |
+
"grad_norm": 3.0801095962524414,
|
| 661 |
+
"learning_rate": 2.3347373080722666e-05,
|
| 662 |
+
"loss": 2.463,
|
| 663 |
+
"step": 4350
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 3.0833917309039944,
|
| 667 |
+
"grad_norm": 2.9642066955566406,
|
| 668 |
+
"learning_rate": 2.315586062556461e-05,
|
| 669 |
+
"loss": 2.4521,
|
| 670 |
+
"step": 4400
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 3.1184302733006306,
|
| 674 |
+
"grad_norm": 3.069679021835327,
|
| 675 |
+
"learning_rate": 2.2962441735562843e-05,
|
| 676 |
+
"loss": 2.466,
|
| 677 |
+
"step": 4450
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 3.1534688156972672,
|
| 681 |
+
"grad_norm": 3.2019903659820557,
|
| 682 |
+
"learning_rate": 2.2767161622440357e-05,
|
| 683 |
+
"loss": 2.473,
|
| 684 |
+
"step": 4500
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 3.1885073580939034,
|
| 688 |
+
"grad_norm": 3.2105560302734375,
|
| 689 |
+
"learning_rate": 2.2570065932981594e-05,
|
| 690 |
+
"loss": 2.4607,
|
| 691 |
+
"step": 4550
|
| 692 |
+
},
|
| 693 |
+
{
|
| 694 |
+
"epoch": 3.2235459004905396,
|
| 695 |
+
"grad_norm": 2.9373726844787598,
|
| 696 |
+
"learning_rate": 2.2371200738362513e-05,
|
| 697 |
+
"loss": 2.4728,
|
| 698 |
+
"step": 4600
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 3.2585844428871757,
|
| 702 |
+
"grad_norm": 3.1778130531311035,
|
| 703 |
+
"learning_rate": 2.2170612523381425e-05,
|
| 704 |
+
"loss": 2.4663,
|
| 705 |
+
"step": 4650
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 3.2936229852838124,
|
| 709 |
+
"grad_norm": 2.7970008850097656,
|
| 710 |
+
"learning_rate": 2.1968348175593148e-05,
|
| 711 |
+
"loss": 2.4619,
|
| 712 |
+
"step": 4700
|
| 713 |
+
},
|
| 714 |
+
{
|
| 715 |
+
"epoch": 3.3286615276804485,
|
| 716 |
+
"grad_norm": 2.9961998462677,
|
| 717 |
+
"learning_rate": 2.1764454974349073e-05,
|
| 718 |
+
"loss": 2.4467,
|
| 719 |
+
"step": 4750
|
| 720 |
+
},
|
| 721 |
+
{
|
| 722 |
+
"epoch": 3.3637000700770847,
|
| 723 |
+
"grad_norm": 2.8396408557891846,
|
| 724 |
+
"learning_rate": 2.155898057974554e-05,
|
| 725 |
+
"loss": 2.467,
|
| 726 |
+
"step": 4800
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 3.3987386124737213,
|
| 730 |
+
"grad_norm": 3.0419580936431885,
|
| 731 |
+
"learning_rate": 2.135197302148332e-05,
|
| 732 |
+
"loss": 2.4724,
|
| 733 |
+
"step": 4850
|
| 734 |
+
},
|
| 735 |
+
{
|
| 736 |
+
"epoch": 3.4337771548703575,
|
| 737 |
+
"grad_norm": 3.01810359954834,
|
| 738 |
+
"learning_rate": 2.1143480687640646e-05,
|
| 739 |
+
"loss": 2.4677,
|
| 740 |
+
"step": 4900
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 3.4688156972669937,
|
| 744 |
+
"grad_norm": 2.8444316387176514,
|
| 745 |
+
"learning_rate": 2.093355231336248e-05,
|
| 746 |
+
"loss": 2.4541,
|
| 747 |
+
"step": 4950
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 3.50385423966363,
|
| 751 |
+
"grad_norm": 3.127967357635498,
|
| 752 |
+
"learning_rate": 2.0722236969468658e-05,
|
| 753 |
+
"loss": 2.4679,
|
| 754 |
+
"step": 5000
|
| 755 |
+
},
|
| 756 |
+
{
|
| 757 |
+
"epoch": 3.50385423966363,
|
| 758 |
+
"eval_bleu": 0.4027446030453843,
|
| 759 |
+
"eval_chrf": 59.07444330732524,
|
| 760 |
+
"eval_loss": 2.5650296211242676,
|
| 761 |
+
"eval_model_preparation_time": 0.0023,
|
| 762 |
+
"eval_runtime": 216.9029,
|
| 763 |
+
"eval_samples_per_second": 99.067,
|
| 764 |
+
"eval_steps_per_second": 0.775,
|
| 765 |
+
"step": 5000
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 3.538892782060266,
|
| 769 |
+
"grad_norm": 3.1810946464538574,
|
| 770 |
+
"learning_rate": 2.0509584050983584e-05,
|
| 771 |
+
"loss": 2.4636,
|
| 772 |
+
"step": 5050
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 3.5739313244569026,
|
| 776 |
+
"grad_norm": 3.019989013671875,
|
| 777 |
+
"learning_rate": 2.0295643265590122e-05,
|
| 778 |
+
"loss": 2.4809,
|
| 779 |
+
"step": 5100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 3.608969866853539,
|
| 783 |
+
"grad_norm": 3.040945291519165,
|
| 784 |
+
"learning_rate": 2.008046462201039e-05,
|
| 785 |
+
"loss": 2.4573,
|
| 786 |
+
"step": 5150
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 3.6440084092501754,
|
| 790 |
+
"grad_norm": 3.132201671600342,
|
| 791 |
+
"learning_rate": 1.9864098418316253e-05,
|
| 792 |
+
"loss": 2.4706,
|
| 793 |
+
"step": 5200
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 3.6790469516468116,
|
| 797 |
+
"grad_norm": 3.0334537029266357,
|
| 798 |
+
"learning_rate": 1.9646595230172085e-05,
|
| 799 |
+
"loss": 2.4756,
|
| 800 |
+
"step": 5250
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 3.714085494043448,
|
| 804 |
+
"grad_norm": 3.1421594619750977,
|
| 805 |
+
"learning_rate": 1.9428005899012718e-05,
|
| 806 |
+
"loss": 2.4515,
|
| 807 |
+
"step": 5300
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 3.749124036440084,
|
| 811 |
+
"grad_norm": 3.2529165744781494,
|
| 812 |
+
"learning_rate": 1.920838152015923e-05,
|
| 813 |
+
"loss": 2.4677,
|
| 814 |
+
"step": 5350
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 3.78416257883672,
|
| 818 |
+
"grad_norm": 3.069092035293579,
|
| 819 |
+
"learning_rate": 1.898777343087542e-05,
|
| 820 |
+
"loss": 2.4742,
|
| 821 |
+
"step": 5400
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 3.8192011212333568,
|
| 825 |
+
"grad_norm": 3.0430819988250732,
|
| 826 |
+
"learning_rate": 1.8766233198367686e-05,
|
| 827 |
+
"loss": 2.4668,
|
| 828 |
+
"step": 5450
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 3.854239663629993,
|
| 832 |
+
"grad_norm": 3.245849370956421,
|
| 833 |
+
"learning_rate": 1.8543812607731164e-05,
|
| 834 |
+
"loss": 2.4758,
|
| 835 |
+
"step": 5500
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 3.8892782060266295,
|
| 839 |
+
"grad_norm": 2.8682169914245605,
|
| 840 |
+
"learning_rate": 1.8320563649844996e-05,
|
| 841 |
+
"loss": 2.4514,
|
| 842 |
+
"step": 5550
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 3.9243167484232657,
|
| 846 |
+
"grad_norm": 2.9359638690948486,
|
| 847 |
+
"learning_rate": 1.8096538509219392e-05,
|
| 848 |
+
"loss": 2.4671,
|
| 849 |
+
"step": 5600
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 3.959355290819902,
|
| 853 |
+
"grad_norm": 2.7711620330810547,
|
| 854 |
+
"learning_rate": 1.7871789551797513e-05,
|
| 855 |
+
"loss": 2.4767,
|
| 856 |
+
"step": 5650
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 3.994393833216538,
|
| 860 |
+
"grad_norm": 3.0792059898376465,
|
| 861 |
+
"learning_rate": 1.764636931271491e-05,
|
| 862 |
+
"loss": 2.4739,
|
| 863 |
+
"step": 5700
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 4.029432375613174,
|
| 867 |
+
"grad_norm": 2.9659736156463623,
|
| 868 |
+
"learning_rate": 1.742033048401943e-05,
|
| 869 |
+
"loss": 2.4126,
|
| 870 |
+
"step": 5750
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 4.06447091800981,
|
| 874 |
+
"grad_norm": 2.9683382511138916,
|
| 875 |
+
"learning_rate": 1.7193725902354417e-05,
|
| 876 |
+
"loss": 2.4032,
|
| 877 |
+
"step": 5800
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 4.0995094604064475,
|
| 881 |
+
"grad_norm": 3.1190121173858643,
|
| 882 |
+
"learning_rate": 1.6966608536608135e-05,
|
| 883 |
+
"loss": 2.3961,
|
| 884 |
+
"step": 5850
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 4.134548002803084,
|
| 888 |
+
"grad_norm": 2.9727401733398438,
|
| 889 |
+
"learning_rate": 1.6739031475532298e-05,
|
| 890 |
+
"loss": 2.3984,
|
| 891 |
+
"step": 5900
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 4.16958654519972,
|
| 895 |
+
"grad_norm": 3.023026704788208,
|
| 896 |
+
"learning_rate": 1.6511047915332517e-05,
|
| 897 |
+
"loss": 2.408,
|
| 898 |
+
"step": 5950
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 4.204625087596356,
|
| 902 |
+
"grad_norm": 3.0228188037872314,
|
| 903 |
+
"learning_rate": 1.6282711147233716e-05,
|
| 904 |
+
"loss": 2.407,
|
| 905 |
+
"step": 6000
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 4.204625087596356,
|
| 909 |
+
"eval_bleu": 0.40780942634660383,
|
| 910 |
+
"eval_chrf": 59.53389218718612,
|
| 911 |
+
"eval_loss": 2.55600643157959,
|
| 912 |
+
"eval_model_preparation_time": 0.0023,
|
| 913 |
+
"eval_runtime": 227.0963,
|
| 914 |
+
"eval_samples_per_second": 94.621,
|
| 915 |
+
"eval_steps_per_second": 0.74,
|
| 916 |
+
"step": 6000
|
| 917 |
+
},
|
| 918 |
+
{
|
| 919 |
+
"epoch": 4.239663629992992,
|
| 920 |
+
"grad_norm": 2.9944727420806885,
|
| 921 |
+
"learning_rate": 1.6054074545023253e-05,
|
| 922 |
+
"loss": 2.4118,
|
| 923 |
+
"step": 6050
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 4.274702172389628,
|
| 927 |
+
"grad_norm": 3.155284881591797,
|
| 928 |
+
"learning_rate": 1.5825191552574802e-05,
|
| 929 |
+
"loss": 2.4142,
|
| 930 |
+
"step": 6100
|
| 931 |
+
},
|
| 932 |
+
{
|
| 933 |
+
"epoch": 4.3097407147862645,
|
| 934 |
+
"grad_norm": 2.874640464782715,
|
| 935 |
+
"learning_rate": 1.5596115671355843e-05,
|
| 936 |
+
"loss": 2.4063,
|
| 937 |
+
"step": 6150
|
| 938 |
+
},
|
| 939 |
+
{
|
| 940 |
+
"epoch": 4.344779257182902,
|
| 941 |
+
"grad_norm": 3.0521326065063477,
|
| 942 |
+
"learning_rate": 1.5366900447921634e-05,
|
| 943 |
+
"loss": 2.4044,
|
| 944 |
+
"step": 6200
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 4.379817799579538,
|
| 948 |
+
"grad_norm": 3.4137041568756104,
|
| 949 |
+
"learning_rate": 1.5137599461398753e-05,
|
| 950 |
+
"loss": 2.4028,
|
| 951 |
+
"step": 6250
|
| 952 |
+
},
|
| 953 |
+
{
|
| 954 |
+
"epoch": 4.414856341976174,
|
| 955 |
+
"grad_norm": 2.841747522354126,
|
| 956 |
+
"learning_rate": 1.4908266310960895e-05,
|
| 957 |
+
"loss": 2.4023,
|
| 958 |
+
"step": 6300
|
| 959 |
+
},
|
| 960 |
+
{
|
| 961 |
+
"epoch": 4.44989488437281,
|
| 962 |
+
"grad_norm": 3.1822855472564697,
|
| 963 |
+
"learning_rate": 1.4678954603300101e-05,
|
| 964 |
+
"loss": 2.4088,
|
| 965 |
+
"step": 6350
|
| 966 |
+
},
|
| 967 |
+
{
|
| 968 |
+
"epoch": 4.484933426769446,
|
| 969 |
+
"grad_norm": 2.915692090988159,
|
| 970 |
+
"learning_rate": 1.4449717940096144e-05,
|
| 971 |
+
"loss": 2.4282,
|
| 972 |
+
"step": 6400
|
| 973 |
+
},
|
| 974 |
+
{
|
| 975 |
+
"epoch": 4.5199719691660825,
|
| 976 |
+
"grad_norm": 3.039735794067383,
|
| 977 |
+
"learning_rate": 1.4220609905487133e-05,
|
| 978 |
+
"loss": 2.414,
|
| 979 |
+
"step": 6450
|
| 980 |
+
},
|
| 981 |
+
{
|
| 982 |
+
"epoch": 4.555010511562719,
|
| 983 |
+
"grad_norm": 3.0917394161224365,
|
| 984 |
+
"learning_rate": 1.399168405354421e-05,
|
| 985 |
+
"loss": 2.4144,
|
| 986 |
+
"step": 6500
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 4.590049053959355,
|
| 990 |
+
"grad_norm": 2.8126635551452637,
|
| 991 |
+
"learning_rate": 1.3762993895753267e-05,
|
| 992 |
+
"loss": 2.4184,
|
| 993 |
+
"step": 6550
|
| 994 |
+
},
|
| 995 |
+
{
|
| 996 |
+
"epoch": 4.625087596355992,
|
| 997 |
+
"grad_norm": 2.917219638824463,
|
| 998 |
+
"learning_rate": 1.3534592888506612e-05,
|
| 999 |
+
"loss": 2.4217,
|
| 1000 |
+
"step": 6600
|
| 1001 |
+
},
|
| 1002 |
+
{
|
| 1003 |
+
"epoch": 4.660126138752628,
|
| 1004 |
+
"grad_norm": 3.1595089435577393,
|
| 1005 |
+
"learning_rate": 1.3306534420607527e-05,
|
| 1006 |
+
"loss": 2.4068,
|
| 1007 |
+
"step": 6650
|
| 1008 |
+
},
|
| 1009 |
+
{
|
| 1010 |
+
"epoch": 4.695164681149264,
|
| 1011 |
+
"grad_norm": 2.9773976802825928,
|
| 1012 |
+
"learning_rate": 1.307887180079063e-05,
|
| 1013 |
+
"loss": 2.3956,
|
| 1014 |
+
"step": 6700
|
| 1015 |
+
},
|
| 1016 |
+
{
|
| 1017 |
+
"epoch": 4.7302032235459,
|
| 1018 |
+
"grad_norm": 3.1548819541931152,
|
| 1019 |
+
"learning_rate": 1.2851658245260907e-05,
|
| 1020 |
+
"loss": 2.4094,
|
| 1021 |
+
"step": 6750
|
| 1022 |
+
},
|
| 1023 |
+
{
|
| 1024 |
+
"epoch": 4.765241765942537,
|
| 1025 |
+
"grad_norm": 3.0027382373809814,
|
| 1026 |
+
"learning_rate": 1.2624946865254472e-05,
|
| 1027 |
+
"loss": 2.3961,
|
| 1028 |
+
"step": 6800
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 4.800280308339173,
|
| 1032 |
+
"grad_norm": 2.9859366416931152,
|
| 1033 |
+
"learning_rate": 1.2398790654623726e-05,
|
| 1034 |
+
"loss": 2.4119,
|
| 1035 |
+
"step": 6850
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 4.83531885073581,
|
| 1039 |
+
"grad_norm": 2.8505470752716064,
|
| 1040 |
+
"learning_rate": 1.2173242477450082e-05,
|
| 1041 |
+
"loss": 2.4102,
|
| 1042 |
+
"step": 6900
|
| 1043 |
+
},
|
| 1044 |
+
{
|
| 1045 |
+
"epoch": 4.870357393132446,
|
| 1046 |
+
"grad_norm": 2.913194179534912,
|
| 1047 |
+
"learning_rate": 1.1948355055686949e-05,
|
| 1048 |
+
"loss": 2.4134,
|
| 1049 |
+
"step": 6950
|
| 1050 |
+
},
|
| 1051 |
+
{
|
| 1052 |
+
"epoch": 4.905395935529082,
|
| 1053 |
+
"grad_norm": 2.9001243114471436,
|
| 1054 |
+
"learning_rate": 1.1724180956835952e-05,
|
| 1055 |
+
"loss": 2.4194,
|
| 1056 |
+
"step": 7000
|
| 1057 |
+
},
|
| 1058 |
+
{
|
| 1059 |
+
"epoch": 4.905395935529082,
|
| 1060 |
+
"eval_bleu": 0.41278158362058287,
|
| 1061 |
+
"eval_chrf": 59.73289041048309,
|
| 1062 |
+
"eval_loss": 2.546309232711792,
|
| 1063 |
+
"eval_model_preparation_time": 0.0023,
|
| 1064 |
+
"eval_runtime": 220.7428,
|
| 1065 |
+
"eval_samples_per_second": 97.344,
|
| 1066 |
+
"eval_steps_per_second": 0.761,
|
| 1067 |
+
"step": 7000
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 4.940434477925718,
|
| 1071 |
+
"grad_norm": 3.1041224002838135,
|
| 1072 |
+
"learning_rate": 1.1500772581659294e-05,
|
| 1073 |
+
"loss": 2.4115,
|
| 1074 |
+
"step": 7050
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 4.9754730203223545,
|
| 1078 |
+
"grad_norm": 2.9533743858337402,
|
| 1079 |
+
"learning_rate": 1.1278182151931026e-05,
|
| 1080 |
+
"loss": 2.4262,
|
| 1081 |
+
"step": 7100
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 5.010511562718991,
|
| 1085 |
+
"grad_norm": 2.8984827995300293,
|
| 1086 |
+
"learning_rate": 1.105646169823021e-05,
|
| 1087 |
+
"loss": 2.3977,
|
| 1088 |
+
"step": 7150
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 5.045550105115627,
|
| 1092 |
+
"grad_norm": 2.8955366611480713,
|
| 1093 |
+
"learning_rate": 1.0835663047778754e-05,
|
| 1094 |
+
"loss": 2.3647,
|
| 1095 |
+
"step": 7200
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 5.080588647512264,
|
| 1099 |
+
"grad_norm": 3.005560874938965,
|
| 1100 |
+
"learning_rate": 1.0615837812326789e-05,
|
| 1101 |
+
"loss": 2.3738,
|
| 1102 |
+
"step": 7250
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 5.1156271899089,
|
| 1106 |
+
"grad_norm": 2.992203712463379,
|
| 1107 |
+
"learning_rate": 1.0397037376088389e-05,
|
| 1108 |
+
"loss": 2.3709,
|
| 1109 |
+
"step": 7300
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 5.150665732305536,
|
| 1113 |
+
"grad_norm": 2.97568416595459,
|
| 1114 |
+
"learning_rate": 1.017931288373052e-05,
|
| 1115 |
+
"loss": 2.3663,
|
| 1116 |
+
"step": 7350
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 5.185704274702172,
|
| 1120 |
+
"grad_norm": 3.005434513092041,
|
| 1121 |
+
"learning_rate": 9.96271522841793e-06,
|
| 1122 |
+
"loss": 2.3586,
|
| 1123 |
+
"step": 7400
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 5.220742817098809,
|
| 1127 |
+
"grad_norm": 3.0041065216064453,
|
| 1128 |
+
"learning_rate": 9.747295039916884e-06,
|
| 1129 |
+
"loss": 2.3762,
|
| 1130 |
+
"step": 7450
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 5.255781359495445,
|
| 1134 |
+
"grad_norm": 2.6837728023529053,
|
| 1135 |
+
"learning_rate": 9.533102672760444e-06,
|
| 1136 |
+
"loss": 2.3622,
|
| 1137 |
+
"step": 7500
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 5.290819901892081,
|
| 1141 |
+
"grad_norm": 3.1430866718292236,
|
| 1142 |
+
"learning_rate": 9.320188194478038e-06,
|
| 1143 |
+
"loss": 2.3706,
|
| 1144 |
+
"step": 7550
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 5.325858444288718,
|
| 1148 |
+
"grad_norm": 2.89998197555542,
|
| 1149 |
+
"learning_rate": 9.10860137389221e-06,
|
| 1150 |
+
"loss": 2.3576,
|
| 1151 |
+
"step": 7600
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 5.360896986685354,
|
| 1155 |
+
"grad_norm": 3.0838561058044434,
|
| 1156 |
+
"learning_rate": 8.898391669485117e-06,
|
| 1157 |
+
"loss": 2.3756,
|
| 1158 |
+
"step": 7650
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 5.39593552908199,
|
| 1162 |
+
"grad_norm": 3.0112173557281494,
|
| 1163 |
+
"learning_rate": 8.689608217837596e-06,
|
| 1164 |
+
"loss": 2.3636,
|
| 1165 |
+
"step": 7700
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 5.4309740714786265,
|
| 1169 |
+
"grad_norm": 3.127743721008301,
|
| 1170 |
+
"learning_rate": 8.482299822143501e-06,
|
| 1171 |
+
"loss": 2.3653,
|
| 1172 |
+
"step": 7750
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 5.466012613875263,
|
| 1176 |
+
"grad_norm": 2.9377152919769287,
|
| 1177 |
+
"learning_rate": 8.276514940801884e-06,
|
| 1178 |
+
"loss": 2.3757,
|
| 1179 |
+
"step": 7800
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 5.501051156271899,
|
| 1183 |
+
"grad_norm": 2.9477999210357666,
|
| 1184 |
+
"learning_rate": 8.072301676089886e-06,
|
| 1185 |
+
"loss": 2.3807,
|
| 1186 |
+
"step": 7850
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 5.536089698668535,
|
| 1190 |
+
"grad_norm": 2.95470929145813,
|
| 1191 |
+
"learning_rate": 7.869707762918785e-06,
|
| 1192 |
+
"loss": 2.3731,
|
| 1193 |
+
"step": 7900
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 5.571128241065171,
|
| 1197 |
+
"grad_norm": 2.960909843444824,
|
| 1198 |
+
"learning_rate": 7.668780557675971e-06,
|
| 1199 |
+
"loss": 2.3607,
|
| 1200 |
+
"step": 7950
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 5.606166783461808,
|
| 1204 |
+
"grad_norm": 3.265627861022949,
|
| 1205 |
+
"learning_rate": 7.469567027155343e-06,
|
| 1206 |
+
"loss": 2.3695,
|
| 1207 |
+
"step": 8000
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 5.606166783461808,
|
| 1211 |
+
"eval_bleu": 0.41580756751861725,
|
| 1212 |
+
"eval_chrf": 60.00173305992757,
|
| 1213 |
+
"eval_loss": 2.5446596145629883,
|
| 1214 |
+
"eval_model_preparation_time": 0.0023,
|
| 1215 |
+
"eval_runtime": 218.3987,
|
| 1216 |
+
"eval_samples_per_second": 98.389,
|
| 1217 |
+
"eval_steps_per_second": 0.769,
|
| 1218 |
+
"step": 8000
|
| 1219 |
+
},
|
| 1220 |
+
{
|
| 1221 |
+
"epoch": 5.6412053258584445,
|
| 1222 |
+
"grad_norm": 2.9601871967315674,
|
| 1223 |
+
"learning_rate": 7.272113737578832e-06,
|
| 1224 |
+
"loss": 2.3608,
|
| 1225 |
+
"step": 8050
|
| 1226 |
+
},
|
| 1227 |
+
{
|
| 1228 |
+
"epoch": 5.676243868255081,
|
| 1229 |
+
"grad_norm": 3.0118155479431152,
|
| 1230 |
+
"learning_rate": 7.0764668437115255e-06,
|
| 1231 |
+
"loss": 2.3711,
|
| 1232 |
+
"step": 8100
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 5.711282410651717,
|
| 1236 |
+
"grad_norm": 2.9687516689300537,
|
| 1237 |
+
"learning_rate": 6.882672078072974e-06,
|
| 1238 |
+
"loss": 2.3756,
|
| 1239 |
+
"step": 8150
|
| 1240 |
+
},
|
| 1241 |
+
{
|
| 1242 |
+
"epoch": 5.746320953048353,
|
| 1243 |
+
"grad_norm": 2.97206449508667,
|
| 1244 |
+
"learning_rate": 6.690774740247211e-06,
|
| 1245 |
+
"loss": 2.3776,
|
| 1246 |
+
"step": 8200
|
| 1247 |
+
},
|
| 1248 |
+
{
|
| 1249 |
+
"epoch": 5.781359495444989,
|
| 1250 |
+
"grad_norm": 3.156719446182251,
|
| 1251 |
+
"learning_rate": 6.500819686293948e-06,
|
| 1252 |
+
"loss": 2.3734,
|
| 1253 |
+
"step": 8250
|
| 1254 |
+
},
|
| 1255 |
+
{
|
| 1256 |
+
"epoch": 5.816398037841626,
|
| 1257 |
+
"grad_norm": 2.8246707916259766,
|
| 1258 |
+
"learning_rate": 6.312851318263469e-06,
|
| 1259 |
+
"loss": 2.3786,
|
| 1260 |
+
"step": 8300
|
| 1261 |
+
},
|
| 1262 |
+
{
|
| 1263 |
+
"epoch": 5.851436580238262,
|
| 1264 |
+
"grad_norm": 2.8077683448791504,
|
| 1265 |
+
"learning_rate": 6.1269135738176305e-06,
|
| 1266 |
+
"loss": 2.3605,
|
| 1267 |
+
"step": 8350
|
| 1268 |
+
},
|
| 1269 |
+
{
|
| 1270 |
+
"epoch": 5.886475122634899,
|
| 1271 |
+
"grad_norm": 3.0974953174591064,
|
| 1272 |
+
"learning_rate": 5.943049915959423e-06,
|
| 1273 |
+
"loss": 2.3819,
|
| 1274 |
+
"step": 8400
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 5.921513665031535,
|
| 1278 |
+
"grad_norm": 3.1544032096862793,
|
| 1279 |
+
"learning_rate": 5.761303322873495e-06,
|
| 1280 |
+
"loss": 2.3774,
|
| 1281 |
+
"step": 8450
|
| 1282 |
+
},
|
| 1283 |
+
{
|
| 1284 |
+
"epoch": 5.956552207428171,
|
| 1285 |
+
"grad_norm": 2.839092493057251,
|
| 1286 |
+
"learning_rate": 5.5817162778799486e-06,
|
| 1287 |
+
"loss": 2.3637,
|
| 1288 |
+
"step": 8500
|
| 1289 |
+
},
|
| 1290 |
+
{
|
| 1291 |
+
"epoch": 5.991590749824807,
|
| 1292 |
+
"grad_norm": 2.850780963897705,
|
| 1293 |
+
"learning_rate": 5.404330759503899e-06,
|
| 1294 |
+
"loss": 2.3702,
|
| 1295 |
+
"step": 8550
|
| 1296 |
+
},
|
| 1297 |
+
{
|
| 1298 |
+
"epoch": 6.026629292221443,
|
| 1299 |
+
"grad_norm": 2.7807981967926025,
|
| 1300 |
+
"learning_rate": 5.229188231662947e-06,
|
| 1301 |
+
"loss": 2.3545,
|
| 1302 |
+
"step": 8600
|
| 1303 |
+
},
|
| 1304 |
+
{
|
| 1305 |
+
"epoch": 6.0616678346180795,
|
| 1306 |
+
"grad_norm": 3.170328378677368,
|
| 1307 |
+
"learning_rate": 5.056329633975009e-06,
|
| 1308 |
+
"loss": 2.3554,
|
| 1309 |
+
"step": 8650
|
| 1310 |
+
},
|
| 1311 |
+
{
|
| 1312 |
+
"epoch": 6.0967063770147165,
|
| 1313 |
+
"grad_norm": 3.196965456008911,
|
| 1314 |
+
"learning_rate": 4.88579537218863e-06,
|
| 1315 |
+
"loss": 2.344,
|
| 1316 |
+
"step": 8700
|
| 1317 |
+
},
|
| 1318 |
+
{
|
| 1319 |
+
"epoch": 6.131744919411353,
|
| 1320 |
+
"grad_norm": 2.923738479614258,
|
| 1321 |
+
"learning_rate": 4.717625308738157e-06,
|
| 1322 |
+
"loss": 2.3456,
|
| 1323 |
+
"step": 8750
|
| 1324 |
+
},
|
| 1325 |
+
{
|
| 1326 |
+
"epoch": 6.166783461807989,
|
| 1327 |
+
"grad_norm": 2.8207976818084717,
|
| 1328 |
+
"learning_rate": 4.551858753425873e-06,
|
| 1329 |
+
"loss": 2.3511,
|
| 1330 |
+
"step": 8800
|
| 1331 |
+
},
|
| 1332 |
+
{
|
| 1333 |
+
"epoch": 6.201822004204625,
|
| 1334 |
+
"grad_norm": 2.9778730869293213,
|
| 1335 |
+
"learning_rate": 4.388534454233318e-06,
|
| 1336 |
+
"loss": 2.3514,
|
| 1337 |
+
"step": 8850
|
| 1338 |
+
},
|
| 1339 |
+
{
|
| 1340 |
+
"epoch": 6.236860546601261,
|
| 1341 |
+
"grad_norm": 2.8051555156707764,
|
| 1342 |
+
"learning_rate": 4.227690588263938e-06,
|
| 1343 |
+
"loss": 2.3506,
|
| 1344 |
+
"step": 8900
|
| 1345 |
+
},
|
| 1346 |
+
{
|
| 1347 |
+
"epoch": 6.271899088997897,
|
| 1348 |
+
"grad_norm": 2.976508617401123,
|
| 1349 |
+
"learning_rate": 4.06936475281918e-06,
|
| 1350 |
+
"loss": 2.3386,
|
| 1351 |
+
"step": 8950
|
| 1352 |
+
},
|
| 1353 |
+
{
|
| 1354 |
+
"epoch": 6.3069376313945344,
|
| 1355 |
+
"grad_norm": 3.1328396797180176,
|
| 1356 |
+
"learning_rate": 3.913593956610104e-06,
|
| 1357 |
+
"loss": 2.3557,
|
| 1358 |
+
"step": 9000
|
| 1359 |
+
},
|
| 1360 |
+
{
|
| 1361 |
+
"epoch": 6.3069376313945344,
|
| 1362 |
+
"eval_bleu": 0.4155468754200415,
|
| 1363 |
+
"eval_chrf": 60.11576195348262,
|
| 1364 |
+
"eval_loss": 2.5433077812194824,
|
| 1365 |
+
"eval_model_preparation_time": 0.0023,
|
| 1366 |
+
"eval_runtime": 220.1119,
|
| 1367 |
+
"eval_samples_per_second": 97.623,
|
| 1368 |
+
"eval_steps_per_second": 0.763,
|
| 1369 |
+
"step": 9000
|
| 1370 |
+
},
|
| 1371 |
+
{
|
| 1372 |
+
"epoch": 6.341976173791171,
|
| 1373 |
+
"grad_norm": 2.8473751544952393,
|
| 1374 |
+
"learning_rate": 3.7604146111065912e-06,
|
| 1375 |
+
"loss": 2.3445,
|
| 1376 |
+
"step": 9050
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 6.377014716187807,
|
| 1380 |
+
"grad_norm": 2.983861207962036,
|
| 1381 |
+
"learning_rate": 3.6098625220261472e-06,
|
| 1382 |
+
"loss": 2.3497,
|
| 1383 |
+
"step": 9100
|
| 1384 |
+
},
|
| 1385 |
+
{
|
| 1386 |
+
"epoch": 6.412053258584443,
|
| 1387 |
+
"grad_norm": 2.9072654247283936,
|
| 1388 |
+
"learning_rate": 3.461972880964326e-06,
|
| 1389 |
+
"loss": 2.3443,
|
| 1390 |
+
"step": 9150
|
| 1391 |
+
},
|
| 1392 |
+
{
|
| 1393 |
+
"epoch": 6.447091800981079,
|
| 1394 |
+
"grad_norm": 3.182084560394287,
|
| 1395 |
+
"learning_rate": 3.3167802571686366e-06,
|
| 1396 |
+
"loss": 2.3488,
|
| 1397 |
+
"step": 9200
|
| 1398 |
+
},
|
| 1399 |
+
{
|
| 1400 |
+
"epoch": 6.482130343377715,
|
| 1401 |
+
"grad_norm": 2.8860056400299072,
|
| 1402 |
+
"learning_rate": 3.1743185894580253e-06,
|
| 1403 |
+
"loss": 2.355,
|
| 1404 |
+
"step": 9250
|
| 1405 |
+
},
|
| 1406 |
+
{
|
| 1407 |
+
"epoch": 6.5171688857743515,
|
| 1408 |
+
"grad_norm": 2.966574192047119,
|
| 1409 |
+
"learning_rate": 3.0346211782896265e-06,
|
| 1410 |
+
"loss": 2.3434,
|
| 1411 |
+
"step": 9300
|
| 1412 |
+
},
|
| 1413 |
+
{
|
| 1414 |
+
"epoch": 6.552207428170988,
|
| 1415 |
+
"grad_norm": 2.6500484943389893,
|
| 1416 |
+
"learning_rate": 2.8977206779748046e-06,
|
| 1417 |
+
"loss": 2.3523,
|
| 1418 |
+
"step": 9350
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 6.587245970567625,
|
| 1422 |
+
"grad_norm": 3.2266886234283447,
|
| 1423 |
+
"learning_rate": 2.763649089046164e-06,
|
| 1424 |
+
"loss": 2.3548,
|
| 1425 |
+
"step": 9400
|
| 1426 |
+
},
|
| 1427 |
+
{
|
| 1428 |
+
"epoch": 6.622284512964261,
|
| 1429 |
+
"grad_norm": 3.055302619934082,
|
| 1430 |
+
"learning_rate": 2.6324377507774415e-06,
|
| 1431 |
+
"loss": 2.3433,
|
| 1432 |
+
"step": 9450
|
| 1433 |
+
},
|
| 1434 |
+
{
|
| 1435 |
+
"epoch": 6.657323055360897,
|
| 1436 |
+
"grad_norm": 3.1511456966400146,
|
| 1437 |
+
"learning_rate": 2.504117333857934e-06,
|
| 1438 |
+
"loss": 2.3462,
|
| 1439 |
+
"step": 9500
|
| 1440 |
+
},
|
| 1441 |
+
{
|
| 1442 |
+
"epoch": 6.692361597757533,
|
| 1443 |
+
"grad_norm": 3.060974359512329,
|
| 1444 |
+
"learning_rate": 2.378717833223199e-06,
|
| 1445 |
+
"loss": 2.3453,
|
| 1446 |
+
"step": 9550
|
| 1447 |
+
},
|
| 1448 |
+
{
|
| 1449 |
+
"epoch": 6.727400140154169,
|
| 1450 |
+
"grad_norm": 2.8586585521698,
|
| 1451 |
+
"learning_rate": 2.256268561043765e-06,
|
| 1452 |
+
"loss": 2.3537,
|
| 1453 |
+
"step": 9600
|
| 1454 |
+
},
|
| 1455 |
+
{
|
| 1456 |
+
"epoch": 6.762438682550806,
|
| 1457 |
+
"grad_norm": 2.7674789428710938,
|
| 1458 |
+
"learning_rate": 2.1367981398733754e-06,
|
| 1459 |
+
"loss": 2.3333,
|
| 1460 |
+
"step": 9650
|
| 1461 |
+
},
|
| 1462 |
+
{
|
| 1463 |
+
"epoch": 6.797477224947443,
|
| 1464 |
+
"grad_norm": 3.034686326980591,
|
| 1465 |
+
"learning_rate": 2.020334495958461e-06,
|
| 1466 |
+
"loss": 2.3304,
|
| 1467 |
+
"step": 9700
|
| 1468 |
+
},
|
| 1469 |
+
{
|
| 1470 |
+
"epoch": 6.832515767344079,
|
| 1471 |
+
"grad_norm": 2.6985130310058594,
|
| 1472 |
+
"learning_rate": 1.9069048527103794e-06,
|
| 1473 |
+
"loss": 2.3431,
|
| 1474 |
+
"step": 9750
|
| 1475 |
+
},
|
| 1476 |
+
{
|
| 1477 |
+
"epoch": 6.867554309740715,
|
| 1478 |
+
"grad_norm": 2.8273324966430664,
|
| 1479 |
+
"learning_rate": 1.7965357243419316e-06,
|
| 1480 |
+
"loss": 2.3273,
|
| 1481 |
+
"step": 9800
|
| 1482 |
+
},
|
| 1483 |
+
{
|
| 1484 |
+
"epoch": 6.902592852137351,
|
| 1485 |
+
"grad_norm": 2.781893730163574,
|
| 1486 |
+
"learning_rate": 1.6892529096696558e-06,
|
| 1487 |
+
"loss": 2.3555,
|
| 1488 |
+
"step": 9850
|
| 1489 |
+
},
|
| 1490 |
+
{
|
| 1491 |
+
"epoch": 6.937631394533987,
|
| 1492 |
+
"grad_norm": 2.8690671920776367,
|
| 1493 |
+
"learning_rate": 1.5850814860833496e-06,
|
| 1494 |
+
"loss": 2.3477,
|
| 1495 |
+
"step": 9900
|
| 1496 |
+
},
|
| 1497 |
+
{
|
| 1498 |
+
"epoch": 6.9726699369306235,
|
| 1499 |
+
"grad_norm": 3.0373213291168213,
|
| 1500 |
+
"learning_rate": 1.4840458036842137e-06,
|
| 1501 |
+
"loss": 2.3502,
|
| 1502 |
+
"step": 9950
|
| 1503 |
+
},
|
| 1504 |
+
{
|
| 1505 |
+
"epoch": 7.00770847932726,
|
| 1506 |
+
"grad_norm": 2.837057590484619,
|
| 1507 |
+
"learning_rate": 1.386169479593017e-06,
|
| 1508 |
+
"loss": 2.3422,
|
| 1509 |
+
"step": 10000
|
| 1510 |
+
},
|
| 1511 |
+
{
|
| 1512 |
+
"epoch": 7.00770847932726,
|
| 1513 |
+
"eval_bleu": 0.41596814878175675,
|
| 1514 |
+
"eval_chrf": 60.135859964059904,
|
| 1515 |
+
"eval_loss": 2.5418286323547363,
|
| 1516 |
+
"eval_model_preparation_time": 0.0023,
|
| 1517 |
+
"eval_runtime": 223.9017,
|
| 1518 |
+
"eval_samples_per_second": 95.971,
|
| 1519 |
+
"eval_steps_per_second": 0.75,
|
| 1520 |
+
"step": 10000
|
| 1521 |
+
},
|
| 1522 |
+
{
|
| 1523 |
+
"epoch": 7.042747021723896,
|
| 1524 |
+
"grad_norm": 2.7915444374084473,
|
| 1525 |
+
"learning_rate": 1.2914753924295964e-06,
|
| 1526 |
+
"loss": 2.3336,
|
| 1527 |
+
"step": 10050
|
| 1528 |
+
},
|
| 1529 |
+
{
|
| 1530 |
+
"epoch": 7.077785564120533,
|
| 1531 |
+
"grad_norm": 3.018704891204834,
|
| 1532 |
+
"learning_rate": 1.1999856769649447e-06,
|
| 1533 |
+
"loss": 2.3328,
|
| 1534 |
+
"step": 10100
|
| 1535 |
+
},
|
| 1536 |
+
{
|
| 1537 |
+
"epoch": 7.112824106517169,
|
| 1538 |
+
"grad_norm": 2.741093873977661,
|
| 1539 |
+
"learning_rate": 1.1117217189472311e-06,
|
| 1540 |
+
"loss": 2.321,
|
| 1541 |
+
"step": 10150
|
| 1542 |
+
},
|
| 1543 |
+
{
|
| 1544 |
+
"epoch": 7.147862648913805,
|
| 1545 |
+
"grad_norm": 2.9243154525756836,
|
| 1546 |
+
"learning_rate": 1.0267041501028578e-06,
|
| 1547 |
+
"loss": 2.3288,
|
| 1548 |
+
"step": 10200
|
| 1549 |
+
},
|
| 1550 |
+
{
|
| 1551 |
+
"epoch": 7.1829011913104415,
|
| 1552 |
+
"grad_norm": 2.7786834239959717,
|
| 1553 |
+
"learning_rate": 9.449528433137883e-07,
|
| 1554 |
+
"loss": 2.3407,
|
| 1555 |
+
"step": 10250
|
| 1556 |
+
},
|
| 1557 |
+
{
|
| 1558 |
+
"epoch": 7.217939733707078,
|
| 1559 |
+
"grad_norm": 2.77376127243042,
|
| 1560 |
+
"learning_rate": 8.664869079722742e-07,
|
| 1561 |
+
"loss": 2.3425,
|
| 1562 |
+
"step": 10300
|
| 1563 |
+
},
|
| 1564 |
+
{
|
| 1565 |
+
"epoch": 7.252978276103714,
|
| 1566 |
+
"grad_norm": 2.7703564167022705,
|
| 1567 |
+
"learning_rate": 7.913246855140044e-07,
|
| 1568 |
+
"loss": 2.3395,
|
| 1569 |
+
"step": 10350
|
| 1570 |
+
},
|
| 1571 |
+
{
|
| 1572 |
+
"epoch": 7.28801681850035,
|
| 1573 |
+
"grad_norm": 2.8868818283081055,
|
| 1574 |
+
"learning_rate": 7.19483745130805e-07,
|
| 1575 |
+
"loss": 2.337,
|
| 1576 |
+
"step": 10400
|
| 1577 |
+
},
|
| 1578 |
+
{
|
| 1579 |
+
"epoch": 7.323055360896987,
|
| 1580 |
+
"grad_norm": 2.7283782958984375,
|
| 1581 |
+
"learning_rate": 6.509808796638272e-07,
|
| 1582 |
+
"loss": 2.3321,
|
| 1583 |
+
"step": 10450
|
| 1584 |
+
},
|
| 1585 |
+
{
|
| 1586 |
+
"epoch": 7.358093903293623,
|
| 1587 |
+
"grad_norm": 2.8174734115600586,
|
| 1588 |
+
"learning_rate": 5.858321016782214e-07,
|
| 1589 |
+
"loss": 2.3305,
|
| 1590 |
+
"step": 10500
|
| 1591 |
+
},
|
| 1592 |
+
{
|
| 1593 |
+
"epoch": 7.393132445690259,
|
| 1594 |
+
"grad_norm": 2.8247244358062744,
|
| 1595 |
+
"learning_rate": 5.240526397201934e-07,
|
| 1596 |
+
"loss": 2.3288,
|
| 1597 |
+
"step": 10550
|
| 1598 |
+
},
|
| 1599 |
+
{
|
| 1600 |
+
"epoch": 7.428170988086896,
|
| 1601 |
+
"grad_norm": 2.5177032947540283,
|
| 1602 |
+
"learning_rate": 4.6565693475732474e-07,
|
| 1603 |
+
"loss": 2.3371,
|
| 1604 |
+
"step": 10600
|
| 1605 |
+
},
|
| 1606 |
+
{
|
| 1607 |
+
"epoch": 7.463209530483532,
|
| 1608 |
+
"grad_norm": 2.9507012367248535,
|
| 1609 |
+
"learning_rate": 4.1065863680299684e-07,
|
| 1610 |
+
"loss": 2.3383,
|
| 1611 |
+
"step": 10650
|
| 1612 |
+
},
|
| 1613 |
+
{
|
| 1614 |
+
"epoch": 7.498248072880168,
|
| 1615 |
+
"grad_norm": 2.915398359298706,
|
| 1616 |
+
"learning_rate": 3.590706017256895e-07,
|
| 1617 |
+
"loss": 2.3261,
|
| 1618 |
+
"step": 10700
|
| 1619 |
+
},
|
| 1620 |
+
{
|
| 1621 |
+
"epoch": 7.533286615276804,
|
| 1622 |
+
"grad_norm": 2.9146273136138916,
|
| 1623 |
+
"learning_rate": 3.10904888243928e-07,
|
| 1624 |
+
"loss": 2.3407,
|
| 1625 |
+
"step": 10750
|
| 1626 |
+
},
|
| 1627 |
+
{
|
| 1628 |
+
"epoch": 7.568325157673441,
|
| 1629 |
+
"grad_norm": 2.9111695289611816,
|
| 1630 |
+
"learning_rate": 2.661727551075432e-07,
|
| 1631 |
+
"loss": 2.3477,
|
| 1632 |
+
"step": 10800
|
| 1633 |
+
},
|
| 1634 |
+
{
|
| 1635 |
+
"epoch": 7.603363700070077,
|
| 1636 |
+
"grad_norm": 2.9273648262023926,
|
| 1637 |
+
"learning_rate": 2.2488465846594475e-07,
|
| 1638 |
+
"loss": 2.3366,
|
| 1639 |
+
"step": 10850
|
| 1640 |
+
},
|
| 1641 |
+
{
|
| 1642 |
+
"epoch": 7.6384022424667135,
|
| 1643 |
+
"grad_norm": 2.8956918716430664,
|
| 1644 |
+
"learning_rate": 1.8705024942398285e-07,
|
| 1645 |
+
"loss": 2.3364,
|
| 1646 |
+
"step": 10900
|
| 1647 |
+
},
|
| 1648 |
+
{
|
| 1649 |
+
"epoch": 7.67344078486335,
|
| 1650 |
+
"grad_norm": 2.6481995582580566,
|
| 1651 |
+
"learning_rate": 1.5267837178600974e-07,
|
| 1652 |
+
"loss": 2.332,
|
| 1653 |
+
"step": 10950
|
| 1654 |
+
},
|
| 1655 |
+
{
|
| 1656 |
+
"epoch": 7.708479327259986,
|
| 1657 |
+
"grad_norm": 2.8763976097106934,
|
| 1658 |
+
"learning_rate": 1.2177705998863308e-07,
|
| 1659 |
+
"loss": 2.3307,
|
| 1660 |
+
"step": 11000
|
| 1661 |
+
},
|
| 1662 |
+
{
|
| 1663 |
+
"epoch": 7.708479327259986,
|
| 1664 |
+
"eval_bleu": 0.4164085587117329,
|
| 1665 |
+
"eval_chrf": 60.169262771025835,
|
| 1666 |
+
"eval_loss": 2.5418813228607178,
|
| 1667 |
+
"eval_model_preparation_time": 0.0023,
|
| 1668 |
+
"eval_runtime": 224.2905,
|
| 1669 |
+
"eval_samples_per_second": 95.804,
|
| 1670 |
+
"eval_steps_per_second": 0.749,
|
| 1671 |
+
"step": 11000
|
| 1672 |
+
}
|
| 1673 |
+
],
|
| 1674 |
+
"logging_steps": 50,
|
| 1675 |
+
"max_steps": 11416,
|
| 1676 |
+
"num_input_tokens_seen": 0,
|
| 1677 |
+
"num_train_epochs": 8,
|
| 1678 |
+
"save_steps": 1000,
|
| 1679 |
+
"stateful_callbacks": {
|
| 1680 |
+
"EarlyStoppingCallback": {
|
| 1681 |
+
"args": {
|
| 1682 |
+
"early_stopping_patience": 5,
|
| 1683 |
+
"early_stopping_threshold": 0.001
|
| 1684 |
+
},
|
| 1685 |
+
"attributes": {
|
| 1686 |
+
"early_stopping_patience_counter": 3
|
| 1687 |
+
}
|
| 1688 |
+
},
|
| 1689 |
+
"TrainerControl": {
|
| 1690 |
+
"args": {
|
| 1691 |
+
"should_epoch_stop": false,
|
| 1692 |
+
"should_evaluate": false,
|
| 1693 |
+
"should_log": false,
|
| 1694 |
+
"should_save": true,
|
| 1695 |
+
"should_training_stop": false
|
| 1696 |
+
},
|
| 1697 |
+
"attributes": {}
|
| 1698 |
+
}
|
| 1699 |
+
},
|
| 1700 |
+
"total_flos": 3.492011858558976e+16,
|
| 1701 |
+
"train_batch_size": 128,
|
| 1702 |
+
"trial_name": null,
|
| 1703 |
+
"trial_params": null
|
| 1704 |
+
}
|
checkpoint-11000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:de315ca70cb9dd86bda39bb42e24958d0b1fa1a5d349f0888fa94ac322554246
|
| 3 |
+
size 5969
|
checkpoint-11000/vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-11416/config.json
ADDED
|
@@ -0,0 +1,54 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"activation_dropout": 0.0,
|
| 3 |
+
"activation_function": "swish",
|
| 4 |
+
"add_bias_logits": false,
|
| 5 |
+
"add_final_layer_norm": false,
|
| 6 |
+
"architectures": [
|
| 7 |
+
"MarianMTModel"
|
| 8 |
+
],
|
| 9 |
+
"attention_dropout": 0.0,
|
| 10 |
+
"classif_dropout": 0.0,
|
| 11 |
+
"classifier_dropout": 0.0,
|
| 12 |
+
"d_model": 512,
|
| 13 |
+
"decoder_attention_heads": 8,
|
| 14 |
+
"decoder_ffn_dim": 2048,
|
| 15 |
+
"decoder_layerdrop": 0.0,
|
| 16 |
+
"decoder_layers": 6,
|
| 17 |
+
"decoder_start_token_id": 64171,
|
| 18 |
+
"decoder_vocab_size": 64172,
|
| 19 |
+
"dropout": 0.1,
|
| 20 |
+
"dtype": "float32",
|
| 21 |
+
"encoder_attention_heads": 8,
|
| 22 |
+
"encoder_ffn_dim": 2048,
|
| 23 |
+
"encoder_layerdrop": 0.0,
|
| 24 |
+
"encoder_layers": 6,
|
| 25 |
+
"eos_token_id": 0,
|
| 26 |
+
"extra_pos_embeddings": 64172,
|
| 27 |
+
"forced_eos_token_id": 0,
|
| 28 |
+
"id2label": {
|
| 29 |
+
"0": "LABEL_0",
|
| 30 |
+
"1": "LABEL_1",
|
| 31 |
+
"2": "LABEL_2"
|
| 32 |
+
},
|
| 33 |
+
"init_std": 0.02,
|
| 34 |
+
"is_encoder_decoder": true,
|
| 35 |
+
"label2id": {
|
| 36 |
+
"LABEL_0": 0,
|
| 37 |
+
"LABEL_1": 1,
|
| 38 |
+
"LABEL_2": 2
|
| 39 |
+
},
|
| 40 |
+
"max_length": null,
|
| 41 |
+
"max_position_embeddings": 512,
|
| 42 |
+
"model_type": "marian",
|
| 43 |
+
"normalize_before": false,
|
| 44 |
+
"normalize_embedding": false,
|
| 45 |
+
"num_beams": null,
|
| 46 |
+
"num_hidden_layers": 6,
|
| 47 |
+
"pad_token_id": 64171,
|
| 48 |
+
"scale_embedding": true,
|
| 49 |
+
"share_encoder_decoder_embeddings": true,
|
| 50 |
+
"static_position_embeddings": true,
|
| 51 |
+
"transformers_version": "4.57.1",
|
| 52 |
+
"use_cache": true,
|
| 53 |
+
"vocab_size": 64172
|
| 54 |
+
}
|
checkpoint-11416/generation_config.json
ADDED
|
@@ -0,0 +1,17 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bad_words_ids": [
|
| 3 |
+
[
|
| 4 |
+
64171
|
| 5 |
+
]
|
| 6 |
+
],
|
| 7 |
+
"decoder_start_token_id": 64171,
|
| 8 |
+
"eos_token_id": [
|
| 9 |
+
0
|
| 10 |
+
],
|
| 11 |
+
"forced_eos_token_id": 0,
|
| 12 |
+
"max_length": 512,
|
| 13 |
+
"num_beams": 6,
|
| 14 |
+
"pad_token_id": 64171,
|
| 15 |
+
"renormalize_logits": true,
|
| 16 |
+
"transformers_version": "4.57.1"
|
| 17 |
+
}
|
checkpoint-11416/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:235a2357396cf9bcfe540413404f8c127873d435dc561626a123e536ada5474c
|
| 3 |
+
size 308263984
|
checkpoint-11416/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d13f0a809e8a26caa759ffe4387f595640e59d76f30339a5579e69c74a1f9ca9
|
| 3 |
+
size 616171979
|
checkpoint-11416/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b0d201d77a46aef2dc02142b7e71a8092a647c6a71d246146442a71ba4442c91
|
| 3 |
+
size 14645
|
checkpoint-11416/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ba65150d860a96f2a7da77e06fb8054f3cee615b5a8b4e6a5f6b2f768948f998
|
| 3 |
+
size 1383
|
checkpoint-11416/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8b890ac7424639a514c6bb26777a5a656a2f3a4853f307b42af53784c5ec2929
|
| 3 |
+
size 1465
|
checkpoint-11416/source.spm
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c4a99ea3602b29fbf901ade8b93a45efa3d7c64eab8fc5fa812383efa327a87d
|
| 3 |
+
size 706917
|
checkpoint-11416/special_tokens_map.json
ADDED
|
@@ -0,0 +1,5 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"eos_token": "</s>",
|
| 3 |
+
"pad_token": "<pad>",
|
| 4 |
+
"unk_token": "<unk>"
|
| 5 |
+
}
|
checkpoint-11416/target.spm
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c6dce5fa58fcd7dde9e81e279b8c075bf42ee558278f73d6fb48e342029d7f19
|
| 3 |
+
size 791194
|
checkpoint-11416/tokenizer_config.json
ADDED
|
@@ -0,0 +1,39 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "</s>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<unk>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"64171": {
|
| 20 |
+
"content": "<pad>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
}
|
| 27 |
+
},
|
| 28 |
+
"clean_up_tokenization_spaces": false,
|
| 29 |
+
"eos_token": "</s>",
|
| 30 |
+
"extra_special_tokens": {},
|
| 31 |
+
"model_max_length": 512,
|
| 32 |
+
"pad_token": "<pad>",
|
| 33 |
+
"separate_vocabs": false,
|
| 34 |
+
"source_lang": "mul",
|
| 35 |
+
"sp_model_kwargs": {},
|
| 36 |
+
"target_lang": "eng",
|
| 37 |
+
"tokenizer_class": "MarianTokenizer",
|
| 38 |
+
"unk_token": "<unk>"
|
| 39 |
+
}
|
checkpoint-11416/trainer_state.json
ADDED
|
@@ -0,0 +1,1760 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 11000,
|
| 3 |
+
"best_metric": 0.4164085587117329,
|
| 4 |
+
"best_model_checkpoint": "models/finetuned-sw-en/checkpoint-11000",
|
| 5 |
+
"epoch": 8.0,
|
| 6 |
+
"eval_steps": 1000,
|
| 7 |
+
"global_step": 11416,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.0350385423966363,
|
| 14 |
+
"grad_norm": 140.77987670898438,
|
| 15 |
+
"learning_rate": 1.2872154115586689e-06,
|
| 16 |
+
"loss": 15.3091,
|
| 17 |
+
"step": 50
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.0700770847932726,
|
| 21 |
+
"grad_norm": 20.989227294921875,
|
| 22 |
+
"learning_rate": 2.6007005253940457e-06,
|
| 23 |
+
"loss": 7.7676,
|
| 24 |
+
"step": 100
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.1051156271899089,
|
| 28 |
+
"grad_norm": 5.4014387130737305,
|
| 29 |
+
"learning_rate": 3.914185639229422e-06,
|
| 30 |
+
"loss": 4.6742,
|
| 31 |
+
"step": 150
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.1401541695865452,
|
| 35 |
+
"grad_norm": 5.136999607086182,
|
| 36 |
+
"learning_rate": 5.227670753064799e-06,
|
| 37 |
+
"loss": 4.2477,
|
| 38 |
+
"step": 200
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.1751927119831815,
|
| 42 |
+
"grad_norm": 5.013916015625,
|
| 43 |
+
"learning_rate": 6.541155866900175e-06,
|
| 44 |
+
"loss": 4.0228,
|
| 45 |
+
"step": 250
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.2102312543798178,
|
| 49 |
+
"grad_norm": 4.574016571044922,
|
| 50 |
+
"learning_rate": 7.854640980735552e-06,
|
| 51 |
+
"loss": 3.8679,
|
| 52 |
+
"step": 300
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.2452697967764541,
|
| 56 |
+
"grad_norm": 4.371525287628174,
|
| 57 |
+
"learning_rate": 9.168126094570928e-06,
|
| 58 |
+
"loss": 3.7254,
|
| 59 |
+
"step": 350
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.2803083391730904,
|
| 63 |
+
"grad_norm": 4.383193492889404,
|
| 64 |
+
"learning_rate": 1.0481611208406305e-05,
|
| 65 |
+
"loss": 3.6216,
|
| 66 |
+
"step": 400
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.3153468815697267,
|
| 70 |
+
"grad_norm": 4.636805534362793,
|
| 71 |
+
"learning_rate": 1.1795096322241681e-05,
|
| 72 |
+
"loss": 3.5232,
|
| 73 |
+
"step": 450
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.350385423966363,
|
| 77 |
+
"grad_norm": 4.319335460662842,
|
| 78 |
+
"learning_rate": 1.3108581436077059e-05,
|
| 79 |
+
"loss": 3.4235,
|
| 80 |
+
"step": 500
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.3854239663629993,
|
| 84 |
+
"grad_norm": 3.9548702239990234,
|
| 85 |
+
"learning_rate": 1.4422066549912435e-05,
|
| 86 |
+
"loss": 3.3257,
|
| 87 |
+
"step": 550
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.4204625087596356,
|
| 91 |
+
"grad_norm": 3.984466314315796,
|
| 92 |
+
"learning_rate": 1.5735551663747813e-05,
|
| 93 |
+
"loss": 3.2771,
|
| 94 |
+
"step": 600
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.4555010511562719,
|
| 98 |
+
"grad_norm": 3.979240655899048,
|
| 99 |
+
"learning_rate": 1.7049036777583185e-05,
|
| 100 |
+
"loss": 3.2041,
|
| 101 |
+
"step": 650
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.4905395935529082,
|
| 105 |
+
"grad_norm": 3.8334100246429443,
|
| 106 |
+
"learning_rate": 1.8362521891418565e-05,
|
| 107 |
+
"loss": 3.1511,
|
| 108 |
+
"step": 700
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.5255781359495445,
|
| 112 |
+
"grad_norm": 3.752436637878418,
|
| 113 |
+
"learning_rate": 1.967600700525394e-05,
|
| 114 |
+
"loss": 3.1209,
|
| 115 |
+
"step": 750
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.5606166783461808,
|
| 119 |
+
"grad_norm": 3.773709535598755,
|
| 120 |
+
"learning_rate": 2.0989492119089317e-05,
|
| 121 |
+
"loss": 3.049,
|
| 122 |
+
"step": 800
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.5956552207428171,
|
| 126 |
+
"grad_norm": 3.9539151191711426,
|
| 127 |
+
"learning_rate": 2.2302977232924692e-05,
|
| 128 |
+
"loss": 3.0337,
|
| 129 |
+
"step": 850
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.6306937631394534,
|
| 133 |
+
"grad_norm": 3.7437591552734375,
|
| 134 |
+
"learning_rate": 2.3616462346760072e-05,
|
| 135 |
+
"loss": 2.9983,
|
| 136 |
+
"step": 900
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.6657323055360898,
|
| 140 |
+
"grad_norm": 3.707627773284912,
|
| 141 |
+
"learning_rate": 2.4929947460595444e-05,
|
| 142 |
+
"loss": 2.9624,
|
| 143 |
+
"step": 950
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.700770847932726,
|
| 147 |
+
"grad_norm": 3.768141508102417,
|
| 148 |
+
"learning_rate": 2.6243432574430824e-05,
|
| 149 |
+
"loss": 2.9167,
|
| 150 |
+
"step": 1000
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.700770847932726,
|
| 154 |
+
"eval_bleu": 0.3189127184480538,
|
| 155 |
+
"eval_chrf": 51.264833545919984,
|
| 156 |
+
"eval_loss": 2.8209688663482666,
|
| 157 |
+
"eval_model_preparation_time": 0.0023,
|
| 158 |
+
"eval_runtime": 221.3909,
|
| 159 |
+
"eval_samples_per_second": 97.059,
|
| 160 |
+
"eval_steps_per_second": 0.759,
|
| 161 |
+
"step": 1000
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 0.7358093903293623,
|
| 165 |
+
"grad_norm": 4.0691142082214355,
|
| 166 |
+
"learning_rate": 2.75569176882662e-05,
|
| 167 |
+
"loss": 2.9229,
|
| 168 |
+
"step": 1050
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 0.7708479327259986,
|
| 172 |
+
"grad_norm": 3.470956325531006,
|
| 173 |
+
"learning_rate": 2.8870402802101576e-05,
|
| 174 |
+
"loss": 2.8988,
|
| 175 |
+
"step": 1100
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 0.8058864751226349,
|
| 179 |
+
"grad_norm": 3.2942214012145996,
|
| 180 |
+
"learning_rate": 2.9999965638050107e-05,
|
| 181 |
+
"loss": 2.9036,
|
| 182 |
+
"step": 1150
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.8409250175192712,
|
| 186 |
+
"grad_norm": 3.5197594165802,
|
| 187 |
+
"learning_rate": 2.999772164915165e-05,
|
| 188 |
+
"loss": 2.8677,
|
| 189 |
+
"step": 1200
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.8759635599159075,
|
| 193 |
+
"grad_norm": 3.3883650302886963,
|
| 194 |
+
"learning_rate": 2.999197193836303e-05,
|
| 195 |
+
"loss": 2.8463,
|
| 196 |
+
"step": 1250
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 0.9110021023125437,
|
| 200 |
+
"grad_norm": 3.5236427783966064,
|
| 201 |
+
"learning_rate": 2.998271784968085e-05,
|
| 202 |
+
"loss": 2.8359,
|
| 203 |
+
"step": 1300
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 0.9460406447091801,
|
| 207 |
+
"grad_norm": 3.7130653858184814,
|
| 208 |
+
"learning_rate": 2.9969961546251093e-05,
|
| 209 |
+
"loss": 2.8076,
|
| 210 |
+
"step": 1350
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 0.9810791871058164,
|
| 214 |
+
"grad_norm": 3.4754154682159424,
|
| 215 |
+
"learning_rate": 2.9953706009863472e-05,
|
| 216 |
+
"loss": 2.8054,
|
| 217 |
+
"step": 1400
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 1.0161177295024526,
|
| 221 |
+
"grad_norm": 3.351656675338745,
|
| 222 |
+
"learning_rate": 2.9933955040254443e-05,
|
| 223 |
+
"loss": 2.7778,
|
| 224 |
+
"step": 1450
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 1.051156271899089,
|
| 228 |
+
"grad_norm": 3.2852399349212646,
|
| 229 |
+
"learning_rate": 2.9910713254219016e-05,
|
| 230 |
+
"loss": 2.7308,
|
| 231 |
+
"step": 1500
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 1.0861948142957254,
|
| 235 |
+
"grad_norm": 3.4650189876556396,
|
| 236 |
+
"learning_rate": 2.9883986084531597e-05,
|
| 237 |
+
"loss": 2.7357,
|
| 238 |
+
"step": 1550
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 1.1212333566923616,
|
| 242 |
+
"grad_norm": 3.4110779762268066,
|
| 243 |
+
"learning_rate": 2.985377977867602e-05,
|
| 244 |
+
"loss": 2.7242,
|
| 245 |
+
"step": 1600
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 1.156271899088998,
|
| 249 |
+
"grad_norm": 3.506265163421631,
|
| 250 |
+
"learning_rate": 2.982010139738526e-05,
|
| 251 |
+
"loss": 2.707,
|
| 252 |
+
"step": 1650
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 1.1913104414856341,
|
| 256 |
+
"grad_norm": 3.2563109397888184,
|
| 257 |
+
"learning_rate": 2.978295881299094e-05,
|
| 258 |
+
"loss": 2.723,
|
| 259 |
+
"step": 1700
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"epoch": 1.2263489838822705,
|
| 263 |
+
"grad_norm": 3.262761354446411,
|
| 264 |
+
"learning_rate": 2.9742360707583205e-05,
|
| 265 |
+
"loss": 2.7156,
|
| 266 |
+
"step": 1750
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 1.2613875262789067,
|
| 270 |
+
"grad_norm": 3.3031880855560303,
|
| 271 |
+
"learning_rate": 2.969831657098125e-05,
|
| 272 |
+
"loss": 2.7032,
|
| 273 |
+
"step": 1800
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 1.296426068675543,
|
| 277 |
+
"grad_norm": 3.6921067237854004,
|
| 278 |
+
"learning_rate": 2.9650836698515077e-05,
|
| 279 |
+
"loss": 2.7066,
|
| 280 |
+
"step": 1850
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 1.3314646110721795,
|
| 284 |
+
"grad_norm": 3.372687578201294,
|
| 285 |
+
"learning_rate": 2.9599932188618993e-05,
|
| 286 |
+
"loss": 2.6926,
|
| 287 |
+
"step": 1900
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 1.3665031534688157,
|
| 291 |
+
"grad_norm": 3.5358870029449463,
|
| 292 |
+
"learning_rate": 2.9545614940237294e-05,
|
| 293 |
+
"loss": 2.6876,
|
| 294 |
+
"step": 1950
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 1.401541695865452,
|
| 298 |
+
"grad_norm": 3.1628262996673584,
|
| 299 |
+
"learning_rate": 2.9487897650042928e-05,
|
| 300 |
+
"loss": 2.6921,
|
| 301 |
+
"step": 2000
|
| 302 |
+
},
|
| 303 |
+
{
|
| 304 |
+
"epoch": 1.401541695865452,
|
| 305 |
+
"eval_bleu": 0.3710693297073462,
|
| 306 |
+
"eval_chrf": 56.162355444042014,
|
| 307 |
+
"eval_loss": 2.6536850929260254,
|
| 308 |
+
"eval_model_preparation_time": 0.0023,
|
| 309 |
+
"eval_runtime": 226.613,
|
| 310 |
+
"eval_samples_per_second": 94.822,
|
| 311 |
+
"eval_steps_per_second": 0.741,
|
| 312 |
+
"step": 2000
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 1.4365802382620882,
|
| 316 |
+
"grad_norm": 3.2792763710021973,
|
| 317 |
+
"learning_rate": 2.942679380946962e-05,
|
| 318 |
+
"loss": 2.6852,
|
| 319 |
+
"step": 2050
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 1.4716187806587246,
|
| 323 |
+
"grad_norm": 3.4099667072296143,
|
| 324 |
+
"learning_rate": 2.9362317701558252e-05,
|
| 325 |
+
"loss": 2.672,
|
| 326 |
+
"step": 2100
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 1.5066573230553608,
|
| 330 |
+
"grad_norm": 3.4033470153808594,
|
| 331 |
+
"learning_rate": 2.9294484397618206e-05,
|
| 332 |
+
"loss": 2.6847,
|
| 333 |
+
"step": 2150
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 1.5416958654519972,
|
| 337 |
+
"grad_norm": 3.6894521713256836,
|
| 338 |
+
"learning_rate": 2.9223309753704418e-05,
|
| 339 |
+
"loss": 2.6789,
|
| 340 |
+
"step": 2200
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 1.5767344078486336,
|
| 344 |
+
"grad_norm": 3.1698200702667236,
|
| 345 |
+
"learning_rate": 2.9148810406911045e-05,
|
| 346 |
+
"loss": 2.6632,
|
| 347 |
+
"step": 2250
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 1.6117729502452698,
|
| 351 |
+
"grad_norm": 3.2145323753356934,
|
| 352 |
+
"learning_rate": 2.9071003771482533e-05,
|
| 353 |
+
"loss": 2.68,
|
| 354 |
+
"step": 2300
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 1.646811492641906,
|
| 358 |
+
"grad_norm": 3.1463518142700195,
|
| 359 |
+
"learning_rate": 2.898990803474302e-05,
|
| 360 |
+
"loss": 2.6609,
|
| 361 |
+
"step": 2350
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 1.6818500350385424,
|
| 365 |
+
"grad_norm": 3.1688807010650635,
|
| 366 |
+
"learning_rate": 2.8905542152845054e-05,
|
| 367 |
+
"loss": 2.6562,
|
| 368 |
+
"step": 2400
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 1.7168885774351788,
|
| 372 |
+
"grad_norm": 3.3895840644836426,
|
| 373 |
+
"learning_rate": 2.881792584633859e-05,
|
| 374 |
+
"loss": 2.6593,
|
| 375 |
+
"step": 2450
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 1.751927119831815,
|
| 379 |
+
"grad_norm": 3.324277639389038,
|
| 380 |
+
"learning_rate": 2.87270795955613e-05,
|
| 381 |
+
"loss": 2.6303,
|
| 382 |
+
"step": 2500
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 1.7869656622284513,
|
| 386 |
+
"grad_norm": 3.1676182746887207,
|
| 387 |
+
"learning_rate": 2.8633024635851283e-05,
|
| 388 |
+
"loss": 2.6473,
|
| 389 |
+
"step": 2550
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 1.8220042046250877,
|
| 393 |
+
"grad_norm": 3.058241128921509,
|
| 394 |
+
"learning_rate": 2.8535782952583308e-05,
|
| 395 |
+
"loss": 2.647,
|
| 396 |
+
"step": 2600
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 1.857042747021724,
|
| 400 |
+
"grad_norm": 3.310572624206543,
|
| 401 |
+
"learning_rate": 2.8435377276029724e-05,
|
| 402 |
+
"loss": 2.6562,
|
| 403 |
+
"step": 2650
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 1.89208128941836,
|
| 407 |
+
"grad_norm": 3.1375229358673096,
|
| 408 |
+
"learning_rate": 2.8331831076047245e-05,
|
| 409 |
+
"loss": 2.6474,
|
| 410 |
+
"step": 2700
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 1.9271198318149965,
|
| 414 |
+
"grad_norm": 3.0486273765563965,
|
| 415 |
+
"learning_rate": 2.822516855659088e-05,
|
| 416 |
+
"loss": 2.6356,
|
| 417 |
+
"step": 2750
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 1.9621583742116329,
|
| 421 |
+
"grad_norm": 3.2743468284606934,
|
| 422 |
+
"learning_rate": 2.8115414650056225e-05,
|
| 423 |
+
"loss": 2.639,
|
| 424 |
+
"step": 2800
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 1.997196916608269,
|
| 428 |
+
"grad_norm": 3.156439781188965,
|
| 429 |
+
"learning_rate": 2.8002595011451535e-05,
|
| 430 |
+
"loss": 2.6246,
|
| 431 |
+
"step": 2850
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 2.032235459004905,
|
| 435 |
+
"grad_norm": 3.180952548980713,
|
| 436 |
+
"learning_rate": 2.7886736012400844e-05,
|
| 437 |
+
"loss": 2.5383,
|
| 438 |
+
"step": 2900
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 2.067274001401542,
|
| 442 |
+
"grad_norm": 3.3652632236480713,
|
| 443 |
+
"learning_rate": 2.7767864734979593e-05,
|
| 444 |
+
"loss": 2.5578,
|
| 445 |
+
"step": 2950
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 2.102312543798178,
|
| 449 |
+
"grad_norm": 3.169829845428467,
|
| 450 |
+
"learning_rate": 2.7646008965384187e-05,
|
| 451 |
+
"loss": 2.5472,
|
| 452 |
+
"step": 3000
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 2.102312543798178,
|
| 456 |
+
"eval_bleu": 0.3885753538883607,
|
| 457 |
+
"eval_chrf": 57.86289492703291,
|
| 458 |
+
"eval_loss": 2.604308605194092,
|
| 459 |
+
"eval_model_preparation_time": 0.0023,
|
| 460 |
+
"eval_runtime": 251.154,
|
| 461 |
+
"eval_samples_per_second": 85.557,
|
| 462 |
+
"eval_steps_per_second": 0.669,
|
| 463 |
+
"step": 3000
|
| 464 |
+
},
|
| 465 |
+
{
|
| 466 |
+
"epoch": 2.137351086194814,
|
| 467 |
+
"grad_norm": 3.187664747238159,
|
| 468 |
+
"learning_rate": 2.7521197187436958e-05,
|
| 469 |
+
"loss": 2.5555,
|
| 470 |
+
"step": 3050
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 2.172389628591451,
|
| 474 |
+
"grad_norm": 3.2646875381469727,
|
| 475 |
+
"learning_rate": 2.7393458575928085e-05,
|
| 476 |
+
"loss": 2.5691,
|
| 477 |
+
"step": 3100
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 2.207428170988087,
|
| 481 |
+
"grad_norm": 3.1656856536865234,
|
| 482 |
+
"learning_rate": 2.7262822989795935e-05,
|
| 483 |
+
"loss": 2.5514,
|
| 484 |
+
"step": 3150
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 2.242466713384723,
|
| 488 |
+
"grad_norm": 2.9856343269348145,
|
| 489 |
+
"learning_rate": 2.712932096514758e-05,
|
| 490 |
+
"loss": 2.5438,
|
| 491 |
+
"step": 3200
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 2.2775052557813593,
|
| 495 |
+
"grad_norm": 3.210883378982544,
|
| 496 |
+
"learning_rate": 2.699298370812093e-05,
|
| 497 |
+
"loss": 2.5334,
|
| 498 |
+
"step": 3250
|
| 499 |
+
},
|
| 500 |
+
{
|
| 501 |
+
"epoch": 2.312543798177996,
|
| 502 |
+
"grad_norm": 3.069847345352173,
|
| 503 |
+
"learning_rate": 2.685384308759031e-05,
|
| 504 |
+
"loss": 2.5514,
|
| 505 |
+
"step": 3300
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 2.347582340574632,
|
| 509 |
+
"grad_norm": 3.252610445022583,
|
| 510 |
+
"learning_rate": 2.6711931627717095e-05,
|
| 511 |
+
"loss": 2.5626,
|
| 512 |
+
"step": 3350
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 2.3826208829712683,
|
| 516 |
+
"grad_norm": 3.202777862548828,
|
| 517 |
+
"learning_rate": 2.6567282500347186e-05,
|
| 518 |
+
"loss": 2.5462,
|
| 519 |
+
"step": 3400
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 2.417659425367905,
|
| 523 |
+
"grad_norm": 3.1640210151672363,
|
| 524 |
+
"learning_rate": 2.641992951725707e-05,
|
| 525 |
+
"loss": 2.5506,
|
| 526 |
+
"step": 3450
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 2.452697967764541,
|
| 530 |
+
"grad_norm": 3.1900179386138916,
|
| 531 |
+
"learning_rate": 2.6269907122250307e-05,
|
| 532 |
+
"loss": 2.555,
|
| 533 |
+
"step": 3500
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 2.4877365101611773,
|
| 537 |
+
"grad_norm": 2.921445846557617,
|
| 538 |
+
"learning_rate": 2.611725038310629e-05,
|
| 539 |
+
"loss": 2.5547,
|
| 540 |
+
"step": 3550
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 2.5227750525578134,
|
| 544 |
+
"grad_norm": 2.9808526039123535,
|
| 545 |
+
"learning_rate": 2.5961994983383144e-05,
|
| 546 |
+
"loss": 2.5642,
|
| 547 |
+
"step": 3600
|
| 548 |
+
},
|
| 549 |
+
{
|
| 550 |
+
"epoch": 2.55781359495445,
|
| 551 |
+
"grad_norm": 3.357646942138672,
|
| 552 |
+
"learning_rate": 2.5804177214076664e-05,
|
| 553 |
+
"loss": 2.5317,
|
| 554 |
+
"step": 3650
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 2.592852137351086,
|
| 558 |
+
"grad_norm": 3.1746153831481934,
|
| 559 |
+
"learning_rate": 2.5643833965137296e-05,
|
| 560 |
+
"loss": 2.5631,
|
| 561 |
+
"step": 3700
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 2.6278906797477224,
|
| 565 |
+
"grad_norm": 3.3090434074401855,
|
| 566 |
+
"learning_rate": 2.548100271684711e-05,
|
| 567 |
+
"loss": 2.5374,
|
| 568 |
+
"step": 3750
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 2.662929222144359,
|
| 572 |
+
"grad_norm": 3.5559844970703125,
|
| 573 |
+
"learning_rate": 2.531572153105876e-05,
|
| 574 |
+
"loss": 2.5469,
|
| 575 |
+
"step": 3800
|
| 576 |
+
},
|
| 577 |
+
{
|
| 578 |
+
"epoch": 2.697967764540995,
|
| 579 |
+
"grad_norm": 3.269394636154175,
|
| 580 |
+
"learning_rate": 2.5148029042298514e-05,
|
| 581 |
+
"loss": 2.5343,
|
| 582 |
+
"step": 3850
|
| 583 |
+
},
|
| 584 |
+
{
|
| 585 |
+
"epoch": 2.7330063069376314,
|
| 586 |
+
"grad_norm": 2.9711074829101562,
|
| 587 |
+
"learning_rate": 2.497796444873543e-05,
|
| 588 |
+
"loss": 2.5388,
|
| 589 |
+
"step": 3900
|
| 590 |
+
},
|
| 591 |
+
{
|
| 592 |
+
"epoch": 2.7680448493342675,
|
| 593 |
+
"grad_norm": 3.074882984161377,
|
| 594 |
+
"learning_rate": 2.480556750301877e-05,
|
| 595 |
+
"loss": 2.5361,
|
| 596 |
+
"step": 3950
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 2.803083391730904,
|
| 600 |
+
"grad_norm": 3.130532741546631,
|
| 601 |
+
"learning_rate": 2.463087850298579e-05,
|
| 602 |
+
"loss": 2.532,
|
| 603 |
+
"step": 4000
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 2.803083391730904,
|
| 607 |
+
"eval_bleu": 0.40072636523788635,
|
| 608 |
+
"eval_chrf": 58.74104751673321,
|
| 609 |
+
"eval_loss": 2.5759472846984863,
|
| 610 |
+
"eval_model_preparation_time": 0.0023,
|
| 611 |
+
"eval_runtime": 225.8985,
|
| 612 |
+
"eval_samples_per_second": 95.122,
|
| 613 |
+
"eval_steps_per_second": 0.744,
|
| 614 |
+
"step": 4000
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 2.8381219341275403,
|
| 618 |
+
"grad_norm": 3.0843958854675293,
|
| 619 |
+
"learning_rate": 2.4453938282242146e-05,
|
| 620 |
+
"loss": 2.547,
|
| 621 |
+
"step": 4050
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 2.8731604765241765,
|
| 625 |
+
"grad_norm": 2.9439711570739746,
|
| 626 |
+
"learning_rate": 2.4274788200616986e-05,
|
| 627 |
+
"loss": 2.5571,
|
| 628 |
+
"step": 4100
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 2.908199018920813,
|
| 632 |
+
"grad_norm": 3.01135516166687,
|
| 633 |
+
"learning_rate": 2.4093470134495104e-05,
|
| 634 |
+
"loss": 2.5365,
|
| 635 |
+
"step": 4150
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"epoch": 2.9432375613174493,
|
| 639 |
+
"grad_norm": 3.141979455947876,
|
| 640 |
+
"learning_rate": 2.3910026467028306e-05,
|
| 641 |
+
"loss": 2.5337,
|
| 642 |
+
"step": 4200
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 2.9782761037140855,
|
| 646 |
+
"grad_norm": 3.1309609413146973,
|
| 647 |
+
"learning_rate": 2.372450007822836e-05,
|
| 648 |
+
"loss": 2.5455,
|
| 649 |
+
"step": 4250
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 3.0133146461107216,
|
| 653 |
+
"grad_norm": 2.8289108276367188,
|
| 654 |
+
"learning_rate": 2.3536934334943747e-05,
|
| 655 |
+
"loss": 2.5126,
|
| 656 |
+
"step": 4300
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 3.0483531885073583,
|
| 660 |
+
"grad_norm": 3.0801095962524414,
|
| 661 |
+
"learning_rate": 2.3347373080722666e-05,
|
| 662 |
+
"loss": 2.463,
|
| 663 |
+
"step": 4350
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 3.0833917309039944,
|
| 667 |
+
"grad_norm": 2.9642066955566406,
|
| 668 |
+
"learning_rate": 2.315586062556461e-05,
|
| 669 |
+
"loss": 2.4521,
|
| 670 |
+
"step": 4400
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 3.1184302733006306,
|
| 674 |
+
"grad_norm": 3.069679021835327,
|
| 675 |
+
"learning_rate": 2.2962441735562843e-05,
|
| 676 |
+
"loss": 2.466,
|
| 677 |
+
"step": 4450
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 3.1534688156972672,
|
| 681 |
+
"grad_norm": 3.2019903659820557,
|
| 682 |
+
"learning_rate": 2.2767161622440357e-05,
|
| 683 |
+
"loss": 2.473,
|
| 684 |
+
"step": 4500
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 3.1885073580939034,
|
| 688 |
+
"grad_norm": 3.2105560302734375,
|
| 689 |
+
"learning_rate": 2.2570065932981594e-05,
|
| 690 |
+
"loss": 2.4607,
|
| 691 |
+
"step": 4550
|
| 692 |
+
},
|
| 693 |
+
{
|
| 694 |
+
"epoch": 3.2235459004905396,
|
| 695 |
+
"grad_norm": 2.9373726844787598,
|
| 696 |
+
"learning_rate": 2.2371200738362513e-05,
|
| 697 |
+
"loss": 2.4728,
|
| 698 |
+
"step": 4600
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 3.2585844428871757,
|
| 702 |
+
"grad_norm": 3.1778130531311035,
|
| 703 |
+
"learning_rate": 2.2170612523381425e-05,
|
| 704 |
+
"loss": 2.4663,
|
| 705 |
+
"step": 4650
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 3.2936229852838124,
|
| 709 |
+
"grad_norm": 2.7970008850097656,
|
| 710 |
+
"learning_rate": 2.1968348175593148e-05,
|
| 711 |
+
"loss": 2.4619,
|
| 712 |
+
"step": 4700
|
| 713 |
+
},
|
| 714 |
+
{
|
| 715 |
+
"epoch": 3.3286615276804485,
|
| 716 |
+
"grad_norm": 2.9961998462677,
|
| 717 |
+
"learning_rate": 2.1764454974349073e-05,
|
| 718 |
+
"loss": 2.4467,
|
| 719 |
+
"step": 4750
|
| 720 |
+
},
|
| 721 |
+
{
|
| 722 |
+
"epoch": 3.3637000700770847,
|
| 723 |
+
"grad_norm": 2.8396408557891846,
|
| 724 |
+
"learning_rate": 2.155898057974554e-05,
|
| 725 |
+
"loss": 2.467,
|
| 726 |
+
"step": 4800
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 3.3987386124737213,
|
| 730 |
+
"grad_norm": 3.0419580936431885,
|
| 731 |
+
"learning_rate": 2.135197302148332e-05,
|
| 732 |
+
"loss": 2.4724,
|
| 733 |
+
"step": 4850
|
| 734 |
+
},
|
| 735 |
+
{
|
| 736 |
+
"epoch": 3.4337771548703575,
|
| 737 |
+
"grad_norm": 3.01810359954834,
|
| 738 |
+
"learning_rate": 2.1143480687640646e-05,
|
| 739 |
+
"loss": 2.4677,
|
| 740 |
+
"step": 4900
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 3.4688156972669937,
|
| 744 |
+
"grad_norm": 2.8444316387176514,
|
| 745 |
+
"learning_rate": 2.093355231336248e-05,
|
| 746 |
+
"loss": 2.4541,
|
| 747 |
+
"step": 4950
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 3.50385423966363,
|
| 751 |
+
"grad_norm": 3.127967357635498,
|
| 752 |
+
"learning_rate": 2.0722236969468658e-05,
|
| 753 |
+
"loss": 2.4679,
|
| 754 |
+
"step": 5000
|
| 755 |
+
},
|
| 756 |
+
{
|
| 757 |
+
"epoch": 3.50385423966363,
|
| 758 |
+
"eval_bleu": 0.4027446030453843,
|
| 759 |
+
"eval_chrf": 59.07444330732524,
|
| 760 |
+
"eval_loss": 2.5650296211242676,
|
| 761 |
+
"eval_model_preparation_time": 0.0023,
|
| 762 |
+
"eval_runtime": 216.9029,
|
| 763 |
+
"eval_samples_per_second": 99.067,
|
| 764 |
+
"eval_steps_per_second": 0.775,
|
| 765 |
+
"step": 5000
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 3.538892782060266,
|
| 769 |
+
"grad_norm": 3.1810946464538574,
|
| 770 |
+
"learning_rate": 2.0509584050983584e-05,
|
| 771 |
+
"loss": 2.4636,
|
| 772 |
+
"step": 5050
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 3.5739313244569026,
|
| 776 |
+
"grad_norm": 3.019989013671875,
|
| 777 |
+
"learning_rate": 2.0295643265590122e-05,
|
| 778 |
+
"loss": 2.4809,
|
| 779 |
+
"step": 5100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 3.608969866853539,
|
| 783 |
+
"grad_norm": 3.040945291519165,
|
| 784 |
+
"learning_rate": 2.008046462201039e-05,
|
| 785 |
+
"loss": 2.4573,
|
| 786 |
+
"step": 5150
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 3.6440084092501754,
|
| 790 |
+
"grad_norm": 3.132201671600342,
|
| 791 |
+
"learning_rate": 1.9864098418316253e-05,
|
| 792 |
+
"loss": 2.4706,
|
| 793 |
+
"step": 5200
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 3.6790469516468116,
|
| 797 |
+
"grad_norm": 3.0334537029266357,
|
| 798 |
+
"learning_rate": 1.9646595230172085e-05,
|
| 799 |
+
"loss": 2.4756,
|
| 800 |
+
"step": 5250
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 3.714085494043448,
|
| 804 |
+
"grad_norm": 3.1421594619750977,
|
| 805 |
+
"learning_rate": 1.9428005899012718e-05,
|
| 806 |
+
"loss": 2.4515,
|
| 807 |
+
"step": 5300
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 3.749124036440084,
|
| 811 |
+
"grad_norm": 3.2529165744781494,
|
| 812 |
+
"learning_rate": 1.920838152015923e-05,
|
| 813 |
+
"loss": 2.4677,
|
| 814 |
+
"step": 5350
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 3.78416257883672,
|
| 818 |
+
"grad_norm": 3.069092035293579,
|
| 819 |
+
"learning_rate": 1.898777343087542e-05,
|
| 820 |
+
"loss": 2.4742,
|
| 821 |
+
"step": 5400
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 3.8192011212333568,
|
| 825 |
+
"grad_norm": 3.0430819988250732,
|
| 826 |
+
"learning_rate": 1.8766233198367686e-05,
|
| 827 |
+
"loss": 2.4668,
|
| 828 |
+
"step": 5450
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 3.854239663629993,
|
| 832 |
+
"grad_norm": 3.245849370956421,
|
| 833 |
+
"learning_rate": 1.8543812607731164e-05,
|
| 834 |
+
"loss": 2.4758,
|
| 835 |
+
"step": 5500
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 3.8892782060266295,
|
| 839 |
+
"grad_norm": 2.8682169914245605,
|
| 840 |
+
"learning_rate": 1.8320563649844996e-05,
|
| 841 |
+
"loss": 2.4514,
|
| 842 |
+
"step": 5550
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 3.9243167484232657,
|
| 846 |
+
"grad_norm": 2.9359638690948486,
|
| 847 |
+
"learning_rate": 1.8096538509219392e-05,
|
| 848 |
+
"loss": 2.4671,
|
| 849 |
+
"step": 5600
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 3.959355290819902,
|
| 853 |
+
"grad_norm": 2.7711620330810547,
|
| 854 |
+
"learning_rate": 1.7871789551797513e-05,
|
| 855 |
+
"loss": 2.4767,
|
| 856 |
+
"step": 5650
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 3.994393833216538,
|
| 860 |
+
"grad_norm": 3.0792059898376465,
|
| 861 |
+
"learning_rate": 1.764636931271491e-05,
|
| 862 |
+
"loss": 2.4739,
|
| 863 |
+
"step": 5700
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 4.029432375613174,
|
| 867 |
+
"grad_norm": 2.9659736156463623,
|
| 868 |
+
"learning_rate": 1.742033048401943e-05,
|
| 869 |
+
"loss": 2.4126,
|
| 870 |
+
"step": 5750
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 4.06447091800981,
|
| 874 |
+
"grad_norm": 2.9683382511138916,
|
| 875 |
+
"learning_rate": 1.7193725902354417e-05,
|
| 876 |
+
"loss": 2.4032,
|
| 877 |
+
"step": 5800
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 4.0995094604064475,
|
| 881 |
+
"grad_norm": 3.1190121173858643,
|
| 882 |
+
"learning_rate": 1.6966608536608135e-05,
|
| 883 |
+
"loss": 2.3961,
|
| 884 |
+
"step": 5850
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 4.134548002803084,
|
| 888 |
+
"grad_norm": 2.9727401733398438,
|
| 889 |
+
"learning_rate": 1.6739031475532298e-05,
|
| 890 |
+
"loss": 2.3984,
|
| 891 |
+
"step": 5900
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 4.16958654519972,
|
| 895 |
+
"grad_norm": 3.023026704788208,
|
| 896 |
+
"learning_rate": 1.6511047915332517e-05,
|
| 897 |
+
"loss": 2.408,
|
| 898 |
+
"step": 5950
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 4.204625087596356,
|
| 902 |
+
"grad_norm": 3.0228188037872314,
|
| 903 |
+
"learning_rate": 1.6282711147233716e-05,
|
| 904 |
+
"loss": 2.407,
|
| 905 |
+
"step": 6000
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 4.204625087596356,
|
| 909 |
+
"eval_bleu": 0.40780942634660383,
|
| 910 |
+
"eval_chrf": 59.53389218718612,
|
| 911 |
+
"eval_loss": 2.55600643157959,
|
| 912 |
+
"eval_model_preparation_time": 0.0023,
|
| 913 |
+
"eval_runtime": 227.0963,
|
| 914 |
+
"eval_samples_per_second": 94.621,
|
| 915 |
+
"eval_steps_per_second": 0.74,
|
| 916 |
+
"step": 6000
|
| 917 |
+
},
|
| 918 |
+
{
|
| 919 |
+
"epoch": 4.239663629992992,
|
| 920 |
+
"grad_norm": 2.9944727420806885,
|
| 921 |
+
"learning_rate": 1.6054074545023253e-05,
|
| 922 |
+
"loss": 2.4118,
|
| 923 |
+
"step": 6050
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 4.274702172389628,
|
| 927 |
+
"grad_norm": 3.155284881591797,
|
| 928 |
+
"learning_rate": 1.5825191552574802e-05,
|
| 929 |
+
"loss": 2.4142,
|
| 930 |
+
"step": 6100
|
| 931 |
+
},
|
| 932 |
+
{
|
| 933 |
+
"epoch": 4.3097407147862645,
|
| 934 |
+
"grad_norm": 2.874640464782715,
|
| 935 |
+
"learning_rate": 1.5596115671355843e-05,
|
| 936 |
+
"loss": 2.4063,
|
| 937 |
+
"step": 6150
|
| 938 |
+
},
|
| 939 |
+
{
|
| 940 |
+
"epoch": 4.344779257182902,
|
| 941 |
+
"grad_norm": 3.0521326065063477,
|
| 942 |
+
"learning_rate": 1.5366900447921634e-05,
|
| 943 |
+
"loss": 2.4044,
|
| 944 |
+
"step": 6200
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 4.379817799579538,
|
| 948 |
+
"grad_norm": 3.4137041568756104,
|
| 949 |
+
"learning_rate": 1.5137599461398753e-05,
|
| 950 |
+
"loss": 2.4028,
|
| 951 |
+
"step": 6250
|
| 952 |
+
},
|
| 953 |
+
{
|
| 954 |
+
"epoch": 4.414856341976174,
|
| 955 |
+
"grad_norm": 2.841747522354126,
|
| 956 |
+
"learning_rate": 1.4908266310960895e-05,
|
| 957 |
+
"loss": 2.4023,
|
| 958 |
+
"step": 6300
|
| 959 |
+
},
|
| 960 |
+
{
|
| 961 |
+
"epoch": 4.44989488437281,
|
| 962 |
+
"grad_norm": 3.1822855472564697,
|
| 963 |
+
"learning_rate": 1.4678954603300101e-05,
|
| 964 |
+
"loss": 2.4088,
|
| 965 |
+
"step": 6350
|
| 966 |
+
},
|
| 967 |
+
{
|
| 968 |
+
"epoch": 4.484933426769446,
|
| 969 |
+
"grad_norm": 2.915692090988159,
|
| 970 |
+
"learning_rate": 1.4449717940096144e-05,
|
| 971 |
+
"loss": 2.4282,
|
| 972 |
+
"step": 6400
|
| 973 |
+
},
|
| 974 |
+
{
|
| 975 |
+
"epoch": 4.5199719691660825,
|
| 976 |
+
"grad_norm": 3.039735794067383,
|
| 977 |
+
"learning_rate": 1.4220609905487133e-05,
|
| 978 |
+
"loss": 2.414,
|
| 979 |
+
"step": 6450
|
| 980 |
+
},
|
| 981 |
+
{
|
| 982 |
+
"epoch": 4.555010511562719,
|
| 983 |
+
"grad_norm": 3.0917394161224365,
|
| 984 |
+
"learning_rate": 1.399168405354421e-05,
|
| 985 |
+
"loss": 2.4144,
|
| 986 |
+
"step": 6500
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 4.590049053959355,
|
| 990 |
+
"grad_norm": 2.8126635551452637,
|
| 991 |
+
"learning_rate": 1.3762993895753267e-05,
|
| 992 |
+
"loss": 2.4184,
|
| 993 |
+
"step": 6550
|
| 994 |
+
},
|
| 995 |
+
{
|
| 996 |
+
"epoch": 4.625087596355992,
|
| 997 |
+
"grad_norm": 2.917219638824463,
|
| 998 |
+
"learning_rate": 1.3534592888506612e-05,
|
| 999 |
+
"loss": 2.4217,
|
| 1000 |
+
"step": 6600
|
| 1001 |
+
},
|
| 1002 |
+
{
|
| 1003 |
+
"epoch": 4.660126138752628,
|
| 1004 |
+
"grad_norm": 3.1595089435577393,
|
| 1005 |
+
"learning_rate": 1.3306534420607527e-05,
|
| 1006 |
+
"loss": 2.4068,
|
| 1007 |
+
"step": 6650
|
| 1008 |
+
},
|
| 1009 |
+
{
|
| 1010 |
+
"epoch": 4.695164681149264,
|
| 1011 |
+
"grad_norm": 2.9773976802825928,
|
| 1012 |
+
"learning_rate": 1.307887180079063e-05,
|
| 1013 |
+
"loss": 2.3956,
|
| 1014 |
+
"step": 6700
|
| 1015 |
+
},
|
| 1016 |
+
{
|
| 1017 |
+
"epoch": 4.7302032235459,
|
| 1018 |
+
"grad_norm": 3.1548819541931152,
|
| 1019 |
+
"learning_rate": 1.2851658245260907e-05,
|
| 1020 |
+
"loss": 2.4094,
|
| 1021 |
+
"step": 6750
|
| 1022 |
+
},
|
| 1023 |
+
{
|
| 1024 |
+
"epoch": 4.765241765942537,
|
| 1025 |
+
"grad_norm": 3.0027382373809814,
|
| 1026 |
+
"learning_rate": 1.2624946865254472e-05,
|
| 1027 |
+
"loss": 2.3961,
|
| 1028 |
+
"step": 6800
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 4.800280308339173,
|
| 1032 |
+
"grad_norm": 2.9859366416931152,
|
| 1033 |
+
"learning_rate": 1.2398790654623726e-05,
|
| 1034 |
+
"loss": 2.4119,
|
| 1035 |
+
"step": 6850
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 4.83531885073581,
|
| 1039 |
+
"grad_norm": 2.8505470752716064,
|
| 1040 |
+
"learning_rate": 1.2173242477450082e-05,
|
| 1041 |
+
"loss": 2.4102,
|
| 1042 |
+
"step": 6900
|
| 1043 |
+
},
|
| 1044 |
+
{
|
| 1045 |
+
"epoch": 4.870357393132446,
|
| 1046 |
+
"grad_norm": 2.913194179534912,
|
| 1047 |
+
"learning_rate": 1.1948355055686949e-05,
|
| 1048 |
+
"loss": 2.4134,
|
| 1049 |
+
"step": 6950
|
| 1050 |
+
},
|
| 1051 |
+
{
|
| 1052 |
+
"epoch": 4.905395935529082,
|
| 1053 |
+
"grad_norm": 2.9001243114471436,
|
| 1054 |
+
"learning_rate": 1.1724180956835952e-05,
|
| 1055 |
+
"loss": 2.4194,
|
| 1056 |
+
"step": 7000
|
| 1057 |
+
},
|
| 1058 |
+
{
|
| 1059 |
+
"epoch": 4.905395935529082,
|
| 1060 |
+
"eval_bleu": 0.41278158362058287,
|
| 1061 |
+
"eval_chrf": 59.73289041048309,
|
| 1062 |
+
"eval_loss": 2.546309232711792,
|
| 1063 |
+
"eval_model_preparation_time": 0.0023,
|
| 1064 |
+
"eval_runtime": 220.7428,
|
| 1065 |
+
"eval_samples_per_second": 97.344,
|
| 1066 |
+
"eval_steps_per_second": 0.761,
|
| 1067 |
+
"step": 7000
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 4.940434477925718,
|
| 1071 |
+
"grad_norm": 3.1041224002838135,
|
| 1072 |
+
"learning_rate": 1.1500772581659294e-05,
|
| 1073 |
+
"loss": 2.4115,
|
| 1074 |
+
"step": 7050
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 4.9754730203223545,
|
| 1078 |
+
"grad_norm": 2.9533743858337402,
|
| 1079 |
+
"learning_rate": 1.1278182151931026e-05,
|
| 1080 |
+
"loss": 2.4262,
|
| 1081 |
+
"step": 7100
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 5.010511562718991,
|
| 1085 |
+
"grad_norm": 2.8984827995300293,
|
| 1086 |
+
"learning_rate": 1.105646169823021e-05,
|
| 1087 |
+
"loss": 2.3977,
|
| 1088 |
+
"step": 7150
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 5.045550105115627,
|
| 1092 |
+
"grad_norm": 2.8955366611480713,
|
| 1093 |
+
"learning_rate": 1.0835663047778754e-05,
|
| 1094 |
+
"loss": 2.3647,
|
| 1095 |
+
"step": 7200
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 5.080588647512264,
|
| 1099 |
+
"grad_norm": 3.005560874938965,
|
| 1100 |
+
"learning_rate": 1.0615837812326789e-05,
|
| 1101 |
+
"loss": 2.3738,
|
| 1102 |
+
"step": 7250
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 5.1156271899089,
|
| 1106 |
+
"grad_norm": 2.992203712463379,
|
| 1107 |
+
"learning_rate": 1.0397037376088389e-05,
|
| 1108 |
+
"loss": 2.3709,
|
| 1109 |
+
"step": 7300
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 5.150665732305536,
|
| 1113 |
+
"grad_norm": 2.97568416595459,
|
| 1114 |
+
"learning_rate": 1.017931288373052e-05,
|
| 1115 |
+
"loss": 2.3663,
|
| 1116 |
+
"step": 7350
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 5.185704274702172,
|
| 1120 |
+
"grad_norm": 3.005434513092041,
|
| 1121 |
+
"learning_rate": 9.96271522841793e-06,
|
| 1122 |
+
"loss": 2.3586,
|
| 1123 |
+
"step": 7400
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 5.220742817098809,
|
| 1127 |
+
"grad_norm": 3.0041065216064453,
|
| 1128 |
+
"learning_rate": 9.747295039916884e-06,
|
| 1129 |
+
"loss": 2.3762,
|
| 1130 |
+
"step": 7450
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 5.255781359495445,
|
| 1134 |
+
"grad_norm": 2.6837728023529053,
|
| 1135 |
+
"learning_rate": 9.533102672760444e-06,
|
| 1136 |
+
"loss": 2.3622,
|
| 1137 |
+
"step": 7500
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 5.290819901892081,
|
| 1141 |
+
"grad_norm": 3.1430866718292236,
|
| 1142 |
+
"learning_rate": 9.320188194478038e-06,
|
| 1143 |
+
"loss": 2.3706,
|
| 1144 |
+
"step": 7550
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 5.325858444288718,
|
| 1148 |
+
"grad_norm": 2.89998197555542,
|
| 1149 |
+
"learning_rate": 9.10860137389221e-06,
|
| 1150 |
+
"loss": 2.3576,
|
| 1151 |
+
"step": 7600
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 5.360896986685354,
|
| 1155 |
+
"grad_norm": 3.0838561058044434,
|
| 1156 |
+
"learning_rate": 8.898391669485117e-06,
|
| 1157 |
+
"loss": 2.3756,
|
| 1158 |
+
"step": 7650
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 5.39593552908199,
|
| 1162 |
+
"grad_norm": 3.0112173557281494,
|
| 1163 |
+
"learning_rate": 8.689608217837596e-06,
|
| 1164 |
+
"loss": 2.3636,
|
| 1165 |
+
"step": 7700
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 5.4309740714786265,
|
| 1169 |
+
"grad_norm": 3.127743721008301,
|
| 1170 |
+
"learning_rate": 8.482299822143501e-06,
|
| 1171 |
+
"loss": 2.3653,
|
| 1172 |
+
"step": 7750
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 5.466012613875263,
|
| 1176 |
+
"grad_norm": 2.9377152919769287,
|
| 1177 |
+
"learning_rate": 8.276514940801884e-06,
|
| 1178 |
+
"loss": 2.3757,
|
| 1179 |
+
"step": 7800
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 5.501051156271899,
|
| 1183 |
+
"grad_norm": 2.9477999210357666,
|
| 1184 |
+
"learning_rate": 8.072301676089886e-06,
|
| 1185 |
+
"loss": 2.3807,
|
| 1186 |
+
"step": 7850
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 5.536089698668535,
|
| 1190 |
+
"grad_norm": 2.95470929145813,
|
| 1191 |
+
"learning_rate": 7.869707762918785e-06,
|
| 1192 |
+
"loss": 2.3731,
|
| 1193 |
+
"step": 7900
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 5.571128241065171,
|
| 1197 |
+
"grad_norm": 2.960909843444824,
|
| 1198 |
+
"learning_rate": 7.668780557675971e-06,
|
| 1199 |
+
"loss": 2.3607,
|
| 1200 |
+
"step": 7950
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 5.606166783461808,
|
| 1204 |
+
"grad_norm": 3.265627861022949,
|
| 1205 |
+
"learning_rate": 7.469567027155343e-06,
|
| 1206 |
+
"loss": 2.3695,
|
| 1207 |
+
"step": 8000
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 5.606166783461808,
|
| 1211 |
+
"eval_bleu": 0.41580756751861725,
|
| 1212 |
+
"eval_chrf": 60.00173305992757,
|
| 1213 |
+
"eval_loss": 2.5446596145629883,
|
| 1214 |
+
"eval_model_preparation_time": 0.0023,
|
| 1215 |
+
"eval_runtime": 218.3987,
|
| 1216 |
+
"eval_samples_per_second": 98.389,
|
| 1217 |
+
"eval_steps_per_second": 0.769,
|
| 1218 |
+
"step": 8000
|
| 1219 |
+
},
|
| 1220 |
+
{
|
| 1221 |
+
"epoch": 5.6412053258584445,
|
| 1222 |
+
"grad_norm": 2.9601871967315674,
|
| 1223 |
+
"learning_rate": 7.272113737578832e-06,
|
| 1224 |
+
"loss": 2.3608,
|
| 1225 |
+
"step": 8050
|
| 1226 |
+
},
|
| 1227 |
+
{
|
| 1228 |
+
"epoch": 5.676243868255081,
|
| 1229 |
+
"grad_norm": 3.0118155479431152,
|
| 1230 |
+
"learning_rate": 7.0764668437115255e-06,
|
| 1231 |
+
"loss": 2.3711,
|
| 1232 |
+
"step": 8100
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 5.711282410651717,
|
| 1236 |
+
"grad_norm": 2.9687516689300537,
|
| 1237 |
+
"learning_rate": 6.882672078072974e-06,
|
| 1238 |
+
"loss": 2.3756,
|
| 1239 |
+
"step": 8150
|
| 1240 |
+
},
|
| 1241 |
+
{
|
| 1242 |
+
"epoch": 5.746320953048353,
|
| 1243 |
+
"grad_norm": 2.97206449508667,
|
| 1244 |
+
"learning_rate": 6.690774740247211e-06,
|
| 1245 |
+
"loss": 2.3776,
|
| 1246 |
+
"step": 8200
|
| 1247 |
+
},
|
| 1248 |
+
{
|
| 1249 |
+
"epoch": 5.781359495444989,
|
| 1250 |
+
"grad_norm": 3.156719446182251,
|
| 1251 |
+
"learning_rate": 6.500819686293948e-06,
|
| 1252 |
+
"loss": 2.3734,
|
| 1253 |
+
"step": 8250
|
| 1254 |
+
},
|
| 1255 |
+
{
|
| 1256 |
+
"epoch": 5.816398037841626,
|
| 1257 |
+
"grad_norm": 2.8246707916259766,
|
| 1258 |
+
"learning_rate": 6.312851318263469e-06,
|
| 1259 |
+
"loss": 2.3786,
|
| 1260 |
+
"step": 8300
|
| 1261 |
+
},
|
| 1262 |
+
{
|
| 1263 |
+
"epoch": 5.851436580238262,
|
| 1264 |
+
"grad_norm": 2.8077683448791504,
|
| 1265 |
+
"learning_rate": 6.1269135738176305e-06,
|
| 1266 |
+
"loss": 2.3605,
|
| 1267 |
+
"step": 8350
|
| 1268 |
+
},
|
| 1269 |
+
{
|
| 1270 |
+
"epoch": 5.886475122634899,
|
| 1271 |
+
"grad_norm": 3.0974953174591064,
|
| 1272 |
+
"learning_rate": 5.943049915959423e-06,
|
| 1273 |
+
"loss": 2.3819,
|
| 1274 |
+
"step": 8400
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 5.921513665031535,
|
| 1278 |
+
"grad_norm": 3.1544032096862793,
|
| 1279 |
+
"learning_rate": 5.761303322873495e-06,
|
| 1280 |
+
"loss": 2.3774,
|
| 1281 |
+
"step": 8450
|
| 1282 |
+
},
|
| 1283 |
+
{
|
| 1284 |
+
"epoch": 5.956552207428171,
|
| 1285 |
+
"grad_norm": 2.839092493057251,
|
| 1286 |
+
"learning_rate": 5.5817162778799486e-06,
|
| 1287 |
+
"loss": 2.3637,
|
| 1288 |
+
"step": 8500
|
| 1289 |
+
},
|
| 1290 |
+
{
|
| 1291 |
+
"epoch": 5.991590749824807,
|
| 1292 |
+
"grad_norm": 2.850780963897705,
|
| 1293 |
+
"learning_rate": 5.404330759503899e-06,
|
| 1294 |
+
"loss": 2.3702,
|
| 1295 |
+
"step": 8550
|
| 1296 |
+
},
|
| 1297 |
+
{
|
| 1298 |
+
"epoch": 6.026629292221443,
|
| 1299 |
+
"grad_norm": 2.7807981967926025,
|
| 1300 |
+
"learning_rate": 5.229188231662947e-06,
|
| 1301 |
+
"loss": 2.3545,
|
| 1302 |
+
"step": 8600
|
| 1303 |
+
},
|
| 1304 |
+
{
|
| 1305 |
+
"epoch": 6.0616678346180795,
|
| 1306 |
+
"grad_norm": 3.170328378677368,
|
| 1307 |
+
"learning_rate": 5.056329633975009e-06,
|
| 1308 |
+
"loss": 2.3554,
|
| 1309 |
+
"step": 8650
|
| 1310 |
+
},
|
| 1311 |
+
{
|
| 1312 |
+
"epoch": 6.0967063770147165,
|
| 1313 |
+
"grad_norm": 3.196965456008911,
|
| 1314 |
+
"learning_rate": 4.88579537218863e-06,
|
| 1315 |
+
"loss": 2.344,
|
| 1316 |
+
"step": 8700
|
| 1317 |
+
},
|
| 1318 |
+
{
|
| 1319 |
+
"epoch": 6.131744919411353,
|
| 1320 |
+
"grad_norm": 2.923738479614258,
|
| 1321 |
+
"learning_rate": 4.717625308738157e-06,
|
| 1322 |
+
"loss": 2.3456,
|
| 1323 |
+
"step": 8750
|
| 1324 |
+
},
|
| 1325 |
+
{
|
| 1326 |
+
"epoch": 6.166783461807989,
|
| 1327 |
+
"grad_norm": 2.8207976818084717,
|
| 1328 |
+
"learning_rate": 4.551858753425873e-06,
|
| 1329 |
+
"loss": 2.3511,
|
| 1330 |
+
"step": 8800
|
| 1331 |
+
},
|
| 1332 |
+
{
|
| 1333 |
+
"epoch": 6.201822004204625,
|
| 1334 |
+
"grad_norm": 2.9778730869293213,
|
| 1335 |
+
"learning_rate": 4.388534454233318e-06,
|
| 1336 |
+
"loss": 2.3514,
|
| 1337 |
+
"step": 8850
|
| 1338 |
+
},
|
| 1339 |
+
{
|
| 1340 |
+
"epoch": 6.236860546601261,
|
| 1341 |
+
"grad_norm": 2.8051555156707764,
|
| 1342 |
+
"learning_rate": 4.227690588263938e-06,
|
| 1343 |
+
"loss": 2.3506,
|
| 1344 |
+
"step": 8900
|
| 1345 |
+
},
|
| 1346 |
+
{
|
| 1347 |
+
"epoch": 6.271899088997897,
|
| 1348 |
+
"grad_norm": 2.976508617401123,
|
| 1349 |
+
"learning_rate": 4.06936475281918e-06,
|
| 1350 |
+
"loss": 2.3386,
|
| 1351 |
+
"step": 8950
|
| 1352 |
+
},
|
| 1353 |
+
{
|
| 1354 |
+
"epoch": 6.3069376313945344,
|
| 1355 |
+
"grad_norm": 3.1328396797180176,
|
| 1356 |
+
"learning_rate": 3.913593956610104e-06,
|
| 1357 |
+
"loss": 2.3557,
|
| 1358 |
+
"step": 9000
|
| 1359 |
+
},
|
| 1360 |
+
{
|
| 1361 |
+
"epoch": 6.3069376313945344,
|
| 1362 |
+
"eval_bleu": 0.4155468754200415,
|
| 1363 |
+
"eval_chrf": 60.11576195348262,
|
| 1364 |
+
"eval_loss": 2.5433077812194824,
|
| 1365 |
+
"eval_model_preparation_time": 0.0023,
|
| 1366 |
+
"eval_runtime": 220.1119,
|
| 1367 |
+
"eval_samples_per_second": 97.623,
|
| 1368 |
+
"eval_steps_per_second": 0.763,
|
| 1369 |
+
"step": 9000
|
| 1370 |
+
},
|
| 1371 |
+
{
|
| 1372 |
+
"epoch": 6.341976173791171,
|
| 1373 |
+
"grad_norm": 2.8473751544952393,
|
| 1374 |
+
"learning_rate": 3.7604146111065912e-06,
|
| 1375 |
+
"loss": 2.3445,
|
| 1376 |
+
"step": 9050
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 6.377014716187807,
|
| 1380 |
+
"grad_norm": 2.983861207962036,
|
| 1381 |
+
"learning_rate": 3.6098625220261472e-06,
|
| 1382 |
+
"loss": 2.3497,
|
| 1383 |
+
"step": 9100
|
| 1384 |
+
},
|
| 1385 |
+
{
|
| 1386 |
+
"epoch": 6.412053258584443,
|
| 1387 |
+
"grad_norm": 2.9072654247283936,
|
| 1388 |
+
"learning_rate": 3.461972880964326e-06,
|
| 1389 |
+
"loss": 2.3443,
|
| 1390 |
+
"step": 9150
|
| 1391 |
+
},
|
| 1392 |
+
{
|
| 1393 |
+
"epoch": 6.447091800981079,
|
| 1394 |
+
"grad_norm": 3.182084560394287,
|
| 1395 |
+
"learning_rate": 3.3167802571686366e-06,
|
| 1396 |
+
"loss": 2.3488,
|
| 1397 |
+
"step": 9200
|
| 1398 |
+
},
|
| 1399 |
+
{
|
| 1400 |
+
"epoch": 6.482130343377715,
|
| 1401 |
+
"grad_norm": 2.8860056400299072,
|
| 1402 |
+
"learning_rate": 3.1743185894580253e-06,
|
| 1403 |
+
"loss": 2.355,
|
| 1404 |
+
"step": 9250
|
| 1405 |
+
},
|
| 1406 |
+
{
|
| 1407 |
+
"epoch": 6.5171688857743515,
|
| 1408 |
+
"grad_norm": 2.966574192047119,
|
| 1409 |
+
"learning_rate": 3.0346211782896265e-06,
|
| 1410 |
+
"loss": 2.3434,
|
| 1411 |
+
"step": 9300
|
| 1412 |
+
},
|
| 1413 |
+
{
|
| 1414 |
+
"epoch": 6.552207428170988,
|
| 1415 |
+
"grad_norm": 2.6500484943389893,
|
| 1416 |
+
"learning_rate": 2.8977206779748046e-06,
|
| 1417 |
+
"loss": 2.3523,
|
| 1418 |
+
"step": 9350
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 6.587245970567625,
|
| 1422 |
+
"grad_norm": 3.2266886234283447,
|
| 1423 |
+
"learning_rate": 2.763649089046164e-06,
|
| 1424 |
+
"loss": 2.3548,
|
| 1425 |
+
"step": 9400
|
| 1426 |
+
},
|
| 1427 |
+
{
|
| 1428 |
+
"epoch": 6.622284512964261,
|
| 1429 |
+
"grad_norm": 3.055302619934082,
|
| 1430 |
+
"learning_rate": 2.6324377507774415e-06,
|
| 1431 |
+
"loss": 2.3433,
|
| 1432 |
+
"step": 9450
|
| 1433 |
+
},
|
| 1434 |
+
{
|
| 1435 |
+
"epoch": 6.657323055360897,
|
| 1436 |
+
"grad_norm": 3.1511456966400146,
|
| 1437 |
+
"learning_rate": 2.504117333857934e-06,
|
| 1438 |
+
"loss": 2.3462,
|
| 1439 |
+
"step": 9500
|
| 1440 |
+
},
|
| 1441 |
+
{
|
| 1442 |
+
"epoch": 6.692361597757533,
|
| 1443 |
+
"grad_norm": 3.060974359512329,
|
| 1444 |
+
"learning_rate": 2.378717833223199e-06,
|
| 1445 |
+
"loss": 2.3453,
|
| 1446 |
+
"step": 9550
|
| 1447 |
+
},
|
| 1448 |
+
{
|
| 1449 |
+
"epoch": 6.727400140154169,
|
| 1450 |
+
"grad_norm": 2.8586585521698,
|
| 1451 |
+
"learning_rate": 2.256268561043765e-06,
|
| 1452 |
+
"loss": 2.3537,
|
| 1453 |
+
"step": 9600
|
| 1454 |
+
},
|
| 1455 |
+
{
|
| 1456 |
+
"epoch": 6.762438682550806,
|
| 1457 |
+
"grad_norm": 2.7674789428710938,
|
| 1458 |
+
"learning_rate": 2.1367981398733754e-06,
|
| 1459 |
+
"loss": 2.3333,
|
| 1460 |
+
"step": 9650
|
| 1461 |
+
},
|
| 1462 |
+
{
|
| 1463 |
+
"epoch": 6.797477224947443,
|
| 1464 |
+
"grad_norm": 3.034686326980591,
|
| 1465 |
+
"learning_rate": 2.020334495958461e-06,
|
| 1466 |
+
"loss": 2.3304,
|
| 1467 |
+
"step": 9700
|
| 1468 |
+
},
|
| 1469 |
+
{
|
| 1470 |
+
"epoch": 6.832515767344079,
|
| 1471 |
+
"grad_norm": 2.6985130310058594,
|
| 1472 |
+
"learning_rate": 1.9069048527103794e-06,
|
| 1473 |
+
"loss": 2.3431,
|
| 1474 |
+
"step": 9750
|
| 1475 |
+
},
|
| 1476 |
+
{
|
| 1477 |
+
"epoch": 6.867554309740715,
|
| 1478 |
+
"grad_norm": 2.8273324966430664,
|
| 1479 |
+
"learning_rate": 1.7965357243419316e-06,
|
| 1480 |
+
"loss": 2.3273,
|
| 1481 |
+
"step": 9800
|
| 1482 |
+
},
|
| 1483 |
+
{
|
| 1484 |
+
"epoch": 6.902592852137351,
|
| 1485 |
+
"grad_norm": 2.781893730163574,
|
| 1486 |
+
"learning_rate": 1.6892529096696558e-06,
|
| 1487 |
+
"loss": 2.3555,
|
| 1488 |
+
"step": 9850
|
| 1489 |
+
},
|
| 1490 |
+
{
|
| 1491 |
+
"epoch": 6.937631394533987,
|
| 1492 |
+
"grad_norm": 2.8690671920776367,
|
| 1493 |
+
"learning_rate": 1.5850814860833496e-06,
|
| 1494 |
+
"loss": 2.3477,
|
| 1495 |
+
"step": 9900
|
| 1496 |
+
},
|
| 1497 |
+
{
|
| 1498 |
+
"epoch": 6.9726699369306235,
|
| 1499 |
+
"grad_norm": 3.0373213291168213,
|
| 1500 |
+
"learning_rate": 1.4840458036842137e-06,
|
| 1501 |
+
"loss": 2.3502,
|
| 1502 |
+
"step": 9950
|
| 1503 |
+
},
|
| 1504 |
+
{
|
| 1505 |
+
"epoch": 7.00770847932726,
|
| 1506 |
+
"grad_norm": 2.837057590484619,
|
| 1507 |
+
"learning_rate": 1.386169479593017e-06,
|
| 1508 |
+
"loss": 2.3422,
|
| 1509 |
+
"step": 10000
|
| 1510 |
+
},
|
| 1511 |
+
{
|
| 1512 |
+
"epoch": 7.00770847932726,
|
| 1513 |
+
"eval_bleu": 0.41596814878175675,
|
| 1514 |
+
"eval_chrf": 60.135859964059904,
|
| 1515 |
+
"eval_loss": 2.5418286323547363,
|
| 1516 |
+
"eval_model_preparation_time": 0.0023,
|
| 1517 |
+
"eval_runtime": 223.9017,
|
| 1518 |
+
"eval_samples_per_second": 95.971,
|
| 1519 |
+
"eval_steps_per_second": 0.75,
|
| 1520 |
+
"step": 10000
|
| 1521 |
+
},
|
| 1522 |
+
{
|
| 1523 |
+
"epoch": 7.042747021723896,
|
| 1524 |
+
"grad_norm": 2.7915444374084473,
|
| 1525 |
+
"learning_rate": 1.2914753924295964e-06,
|
| 1526 |
+
"loss": 2.3336,
|
| 1527 |
+
"step": 10050
|
| 1528 |
+
},
|
| 1529 |
+
{
|
| 1530 |
+
"epoch": 7.077785564120533,
|
| 1531 |
+
"grad_norm": 3.018704891204834,
|
| 1532 |
+
"learning_rate": 1.1999856769649447e-06,
|
| 1533 |
+
"loss": 2.3328,
|
| 1534 |
+
"step": 10100
|
| 1535 |
+
},
|
| 1536 |
+
{
|
| 1537 |
+
"epoch": 7.112824106517169,
|
| 1538 |
+
"grad_norm": 2.741093873977661,
|
| 1539 |
+
"learning_rate": 1.1117217189472311e-06,
|
| 1540 |
+
"loss": 2.321,
|
| 1541 |
+
"step": 10150
|
| 1542 |
+
},
|
| 1543 |
+
{
|
| 1544 |
+
"epoch": 7.147862648913805,
|
| 1545 |
+
"grad_norm": 2.9243154525756836,
|
| 1546 |
+
"learning_rate": 1.0267041501028578e-06,
|
| 1547 |
+
"loss": 2.3288,
|
| 1548 |
+
"step": 10200
|
| 1549 |
+
},
|
| 1550 |
+
{
|
| 1551 |
+
"epoch": 7.1829011913104415,
|
| 1552 |
+
"grad_norm": 2.7786834239959717,
|
| 1553 |
+
"learning_rate": 9.449528433137883e-07,
|
| 1554 |
+
"loss": 2.3407,
|
| 1555 |
+
"step": 10250
|
| 1556 |
+
},
|
| 1557 |
+
{
|
| 1558 |
+
"epoch": 7.217939733707078,
|
| 1559 |
+
"grad_norm": 2.77376127243042,
|
| 1560 |
+
"learning_rate": 8.664869079722742e-07,
|
| 1561 |
+
"loss": 2.3425,
|
| 1562 |
+
"step": 10300
|
| 1563 |
+
},
|
| 1564 |
+
{
|
| 1565 |
+
"epoch": 7.252978276103714,
|
| 1566 |
+
"grad_norm": 2.7703564167022705,
|
| 1567 |
+
"learning_rate": 7.913246855140044e-07,
|
| 1568 |
+
"loss": 2.3395,
|
| 1569 |
+
"step": 10350
|
| 1570 |
+
},
|
| 1571 |
+
{
|
| 1572 |
+
"epoch": 7.28801681850035,
|
| 1573 |
+
"grad_norm": 2.8868818283081055,
|
| 1574 |
+
"learning_rate": 7.19483745130805e-07,
|
| 1575 |
+
"loss": 2.337,
|
| 1576 |
+
"step": 10400
|
| 1577 |
+
},
|
| 1578 |
+
{
|
| 1579 |
+
"epoch": 7.323055360896987,
|
| 1580 |
+
"grad_norm": 2.7283782958984375,
|
| 1581 |
+
"learning_rate": 6.509808796638272e-07,
|
| 1582 |
+
"loss": 2.3321,
|
| 1583 |
+
"step": 10450
|
| 1584 |
+
},
|
| 1585 |
+
{
|
| 1586 |
+
"epoch": 7.358093903293623,
|
| 1587 |
+
"grad_norm": 2.8174734115600586,
|
| 1588 |
+
"learning_rate": 5.858321016782214e-07,
|
| 1589 |
+
"loss": 2.3305,
|
| 1590 |
+
"step": 10500
|
| 1591 |
+
},
|
| 1592 |
+
{
|
| 1593 |
+
"epoch": 7.393132445690259,
|
| 1594 |
+
"grad_norm": 2.8247244358062744,
|
| 1595 |
+
"learning_rate": 5.240526397201934e-07,
|
| 1596 |
+
"loss": 2.3288,
|
| 1597 |
+
"step": 10550
|
| 1598 |
+
},
|
| 1599 |
+
{
|
| 1600 |
+
"epoch": 7.428170988086896,
|
| 1601 |
+
"grad_norm": 2.5177032947540283,
|
| 1602 |
+
"learning_rate": 4.6565693475732474e-07,
|
| 1603 |
+
"loss": 2.3371,
|
| 1604 |
+
"step": 10600
|
| 1605 |
+
},
|
| 1606 |
+
{
|
| 1607 |
+
"epoch": 7.463209530483532,
|
| 1608 |
+
"grad_norm": 2.9507012367248535,
|
| 1609 |
+
"learning_rate": 4.1065863680299684e-07,
|
| 1610 |
+
"loss": 2.3383,
|
| 1611 |
+
"step": 10650
|
| 1612 |
+
},
|
| 1613 |
+
{
|
| 1614 |
+
"epoch": 7.498248072880168,
|
| 1615 |
+
"grad_norm": 2.915398359298706,
|
| 1616 |
+
"learning_rate": 3.590706017256895e-07,
|
| 1617 |
+
"loss": 2.3261,
|
| 1618 |
+
"step": 10700
|
| 1619 |
+
},
|
| 1620 |
+
{
|
| 1621 |
+
"epoch": 7.533286615276804,
|
| 1622 |
+
"grad_norm": 2.9146273136138916,
|
| 1623 |
+
"learning_rate": 3.10904888243928e-07,
|
| 1624 |
+
"loss": 2.3407,
|
| 1625 |
+
"step": 10750
|
| 1626 |
+
},
|
| 1627 |
+
{
|
| 1628 |
+
"epoch": 7.568325157673441,
|
| 1629 |
+
"grad_norm": 2.9111695289611816,
|
| 1630 |
+
"learning_rate": 2.661727551075432e-07,
|
| 1631 |
+
"loss": 2.3477,
|
| 1632 |
+
"step": 10800
|
| 1633 |
+
},
|
| 1634 |
+
{
|
| 1635 |
+
"epoch": 7.603363700070077,
|
| 1636 |
+
"grad_norm": 2.9273648262023926,
|
| 1637 |
+
"learning_rate": 2.2488465846594475e-07,
|
| 1638 |
+
"loss": 2.3366,
|
| 1639 |
+
"step": 10850
|
| 1640 |
+
},
|
| 1641 |
+
{
|
| 1642 |
+
"epoch": 7.6384022424667135,
|
| 1643 |
+
"grad_norm": 2.8956918716430664,
|
| 1644 |
+
"learning_rate": 1.8705024942398285e-07,
|
| 1645 |
+
"loss": 2.3364,
|
| 1646 |
+
"step": 10900
|
| 1647 |
+
},
|
| 1648 |
+
{
|
| 1649 |
+
"epoch": 7.67344078486335,
|
| 1650 |
+
"grad_norm": 2.6481995582580566,
|
| 1651 |
+
"learning_rate": 1.5267837178600974e-07,
|
| 1652 |
+
"loss": 2.332,
|
| 1653 |
+
"step": 10950
|
| 1654 |
+
},
|
| 1655 |
+
{
|
| 1656 |
+
"epoch": 7.708479327259986,
|
| 1657 |
+
"grad_norm": 2.8763976097106934,
|
| 1658 |
+
"learning_rate": 1.2177705998863308e-07,
|
| 1659 |
+
"loss": 2.3307,
|
| 1660 |
+
"step": 11000
|
| 1661 |
+
},
|
| 1662 |
+
{
|
| 1663 |
+
"epoch": 7.708479327259986,
|
| 1664 |
+
"eval_bleu": 0.4164085587117329,
|
| 1665 |
+
"eval_chrf": 60.169262771025835,
|
| 1666 |
+
"eval_loss": 2.5418813228607178,
|
| 1667 |
+
"eval_model_preparation_time": 0.0023,
|
| 1668 |
+
"eval_runtime": 224.2905,
|
| 1669 |
+
"eval_samples_per_second": 95.804,
|
| 1670 |
+
"eval_steps_per_second": 0.749,
|
| 1671 |
+
"step": 11000
|
| 1672 |
+
},
|
| 1673 |
+
{
|
| 1674 |
+
"epoch": 7.743517869656622,
|
| 1675 |
+
"grad_norm": 2.7782721519470215,
|
| 1676 |
+
"learning_rate": 9.435353722266338e-08,
|
| 1677 |
+
"loss": 2.3311,
|
| 1678 |
+
"step": 11050
|
| 1679 |
+
},
|
| 1680 |
+
{
|
| 1681 |
+
"epoch": 7.778556412053258,
|
| 1682 |
+
"grad_norm": 2.9786736965179443,
|
| 1683 |
+
"learning_rate": 7.041421374469281e-08,
|
| 1684 |
+
"loss": 2.3564,
|
| 1685 |
+
"step": 11100
|
| 1686 |
+
},
|
| 1687 |
+
{
|
| 1688 |
+
"epoch": 7.813594954449895,
|
| 1689 |
+
"grad_norm": 3.7628660202026367,
|
| 1690 |
+
"learning_rate": 4.9964685378695586e-08,
|
| 1691 |
+
"loss": 2.3378,
|
| 1692 |
+
"step": 11150
|
| 1693 |
+
},
|
| 1694 |
+
{
|
| 1695 |
+
"epoch": 7.848633496846531,
|
| 1696 |
+
"grad_norm": 3.0894999504089355,
|
| 1697 |
+
"learning_rate": 3.3009732208000855e-08,
|
| 1698 |
+
"loss": 2.3368,
|
| 1699 |
+
"step": 11200
|
| 1700 |
+
},
|
| 1701 |
+
{
|
| 1702 |
+
"epoch": 7.883672039243168,
|
| 1703 |
+
"grad_norm": 2.8265011310577393,
|
| 1704 |
+
"learning_rate": 1.9553317457948196e-08,
|
| 1705 |
+
"loss": 2.3449,
|
| 1706 |
+
"step": 11250
|
| 1707 |
+
},
|
| 1708 |
+
{
|
| 1709 |
+
"epoch": 7.918710581639804,
|
| 1710 |
+
"grad_norm": 2.8764238357543945,
|
| 1711 |
+
"learning_rate": 9.598586569481981e-09,
|
| 1712 |
+
"loss": 2.3345,
|
| 1713 |
+
"step": 11300
|
| 1714 |
+
},
|
| 1715 |
+
{
|
| 1716 |
+
"epoch": 7.95374912403644,
|
| 1717 |
+
"grad_norm": 3.020198345184326,
|
| 1718 |
+
"learning_rate": 3.14786646390286e-09,
|
| 1719 |
+
"loss": 2.3379,
|
| 1720 |
+
"step": 11350
|
| 1721 |
+
},
|
| 1722 |
+
{
|
| 1723 |
+
"epoch": 7.988787666433076,
|
| 1724 |
+
"grad_norm": 2.8651113510131836,
|
| 1725 |
+
"learning_rate": 2.0266499895449374e-10,
|
| 1726 |
+
"loss": 2.3303,
|
| 1727 |
+
"step": 11400
|
| 1728 |
+
}
|
| 1729 |
+
],
|
| 1730 |
+
"logging_steps": 50,
|
| 1731 |
+
"max_steps": 11416,
|
| 1732 |
+
"num_input_tokens_seen": 0,
|
| 1733 |
+
"num_train_epochs": 8,
|
| 1734 |
+
"save_steps": 1000,
|
| 1735 |
+
"stateful_callbacks": {
|
| 1736 |
+
"EarlyStoppingCallback": {
|
| 1737 |
+
"args": {
|
| 1738 |
+
"early_stopping_patience": 5,
|
| 1739 |
+
"early_stopping_threshold": 0.001
|
| 1740 |
+
},
|
| 1741 |
+
"attributes": {
|
| 1742 |
+
"early_stopping_patience_counter": 3
|
| 1743 |
+
}
|
| 1744 |
+
},
|
| 1745 |
+
"TrainerControl": {
|
| 1746 |
+
"args": {
|
| 1747 |
+
"should_epoch_stop": false,
|
| 1748 |
+
"should_evaluate": false,
|
| 1749 |
+
"should_log": false,
|
| 1750 |
+
"should_save": true,
|
| 1751 |
+
"should_training_stop": true
|
| 1752 |
+
},
|
| 1753 |
+
"attributes": {}
|
| 1754 |
+
}
|
| 1755 |
+
},
|
| 1756 |
+
"total_flos": 3.623867864639078e+16,
|
| 1757 |
+
"train_batch_size": 128,
|
| 1758 |
+
"trial_name": null,
|
| 1759 |
+
"trial_params": null
|
| 1760 |
+
}
|
checkpoint-11416/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:de315ca70cb9dd86bda39bb42e24958d0b1fa1a5d349f0888fa94ac322554246
|
| 3 |
+
size 5969
|
checkpoint-11416/vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-384/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5969
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:de315ca70cb9dd86bda39bb42e24958d0b1fa1a5d349f0888fa94ac322554246
|
| 3 |
size 5969
|
emergency_checkpoint/config.json
ADDED
|
@@ -0,0 +1,54 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"activation_dropout": 0.0,
|
| 3 |
+
"activation_function": "swish",
|
| 4 |
+
"add_bias_logits": false,
|
| 5 |
+
"add_final_layer_norm": false,
|
| 6 |
+
"architectures": [
|
| 7 |
+
"MarianMTModel"
|
| 8 |
+
],
|
| 9 |
+
"attention_dropout": 0.0,
|
| 10 |
+
"classif_dropout": 0.0,
|
| 11 |
+
"classifier_dropout": 0.0,
|
| 12 |
+
"d_model": 512,
|
| 13 |
+
"decoder_attention_heads": 8,
|
| 14 |
+
"decoder_ffn_dim": 2048,
|
| 15 |
+
"decoder_layerdrop": 0.0,
|
| 16 |
+
"decoder_layers": 6,
|
| 17 |
+
"decoder_start_token_id": 64171,
|
| 18 |
+
"decoder_vocab_size": 64172,
|
| 19 |
+
"dropout": 0.1,
|
| 20 |
+
"dtype": "float32",
|
| 21 |
+
"encoder_attention_heads": 8,
|
| 22 |
+
"encoder_ffn_dim": 2048,
|
| 23 |
+
"encoder_layerdrop": 0.0,
|
| 24 |
+
"encoder_layers": 6,
|
| 25 |
+
"eos_token_id": 0,
|
| 26 |
+
"extra_pos_embeddings": 64172,
|
| 27 |
+
"forced_eos_token_id": 0,
|
| 28 |
+
"id2label": {
|
| 29 |
+
"0": "LABEL_0",
|
| 30 |
+
"1": "LABEL_1",
|
| 31 |
+
"2": "LABEL_2"
|
| 32 |
+
},
|
| 33 |
+
"init_std": 0.02,
|
| 34 |
+
"is_encoder_decoder": true,
|
| 35 |
+
"label2id": {
|
| 36 |
+
"LABEL_0": 0,
|
| 37 |
+
"LABEL_1": 1,
|
| 38 |
+
"LABEL_2": 2
|
| 39 |
+
},
|
| 40 |
+
"max_length": null,
|
| 41 |
+
"max_position_embeddings": 512,
|
| 42 |
+
"model_type": "marian",
|
| 43 |
+
"normalize_before": false,
|
| 44 |
+
"normalize_embedding": false,
|
| 45 |
+
"num_beams": null,
|
| 46 |
+
"num_hidden_layers": 6,
|
| 47 |
+
"pad_token_id": 64171,
|
| 48 |
+
"scale_embedding": true,
|
| 49 |
+
"share_encoder_decoder_embeddings": true,
|
| 50 |
+
"static_position_embeddings": true,
|
| 51 |
+
"transformers_version": "4.57.1",
|
| 52 |
+
"use_cache": true,
|
| 53 |
+
"vocab_size": 64172
|
| 54 |
+
}
|
emergency_checkpoint/generation_config.json
ADDED
|
@@ -0,0 +1,17 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bad_words_ids": [
|
| 3 |
+
[
|
| 4 |
+
64171
|
| 5 |
+
]
|
| 6 |
+
],
|
| 7 |
+
"decoder_start_token_id": 64171,
|
| 8 |
+
"eos_token_id": [
|
| 9 |
+
0
|
| 10 |
+
],
|
| 11 |
+
"forced_eos_token_id": 0,
|
| 12 |
+
"max_length": 512,
|
| 13 |
+
"num_beams": 6,
|
| 14 |
+
"pad_token_id": 64171,
|
| 15 |
+
"renormalize_logits": true,
|
| 16 |
+
"transformers_version": "4.57.1"
|
| 17 |
+
}
|
emergency_checkpoint/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:968e4e015c98e23af3da62d2661b4ed129a2b3a040b7e9d159ec6b6e5b3b4def
|
| 3 |
+
size 308263984
|
emergency_checkpoint/source.spm
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c4a99ea3602b29fbf901ade8b93a45efa3d7c64eab8fc5fa812383efa327a87d
|
| 3 |
+
size 706917
|
emergency_checkpoint/special_tokens_map.json
ADDED
|
@@ -0,0 +1,5 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"eos_token": "</s>",
|
| 3 |
+
"pad_token": "<pad>",
|
| 4 |
+
"unk_token": "<unk>"
|
| 5 |
+
}
|
emergency_checkpoint/target.spm
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c6dce5fa58fcd7dde9e81e279b8c075bf42ee558278f73d6fb48e342029d7f19
|
| 3 |
+
size 791194
|
emergency_checkpoint/tokenizer_config.json
ADDED
|
@@ -0,0 +1,39 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "</s>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<unk>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"64171": {
|
| 20 |
+
"content": "<pad>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
}
|
| 27 |
+
},
|
| 28 |
+
"clean_up_tokenization_spaces": false,
|
| 29 |
+
"eos_token": "</s>",
|
| 30 |
+
"extra_special_tokens": {},
|
| 31 |
+
"model_max_length": 512,
|
| 32 |
+
"pad_token": "<pad>",
|
| 33 |
+
"separate_vocabs": false,
|
| 34 |
+
"source_lang": "mul",
|
| 35 |
+
"sp_model_kwargs": {},
|
| 36 |
+
"target_lang": "eng",
|
| 37 |
+
"tokenizer_class": "MarianTokenizer",
|
| 38 |
+
"unk_token": "<unk>"
|
| 39 |
+
}
|
emergency_checkpoint/vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
test_results.json
CHANGED
|
@@ -3,9 +3,9 @@
|
|
| 3 |
"eval_model_preparation_time": 0.0023,
|
| 4 |
"eval_bleu": 0.23200950322593514,
|
| 5 |
"eval_chrf": 43.01447112724294,
|
| 6 |
-
"eval_runtime":
|
| 7 |
-
"eval_samples_per_second":
|
| 8 |
-
"eval_steps_per_second": 0.
|
| 9 |
"epoch": 8.0,
|
| 10 |
"eval_comet": 0.16610798774550453,
|
| 11 |
"eval_comet_std": 0.24794758031968306,
|
|
|
|
| 3 |
"eval_model_preparation_time": 0.0023,
|
| 4 |
"eval_bleu": 0.23200950322593514,
|
| 5 |
"eval_chrf": 43.01447112724294,
|
| 6 |
+
"eval_runtime": 3.5958,
|
| 7 |
+
"eval_samples_per_second": 100.117,
|
| 8 |
+
"eval_steps_per_second": 0.834,
|
| 9 |
"epoch": 8.0,
|
| 10 |
"eval_comet": 0.16610798774550453,
|
| 11 |
"eval_comet_std": 0.24794758031968306,
|
train_results.json
CHANGED
|
@@ -1,7 +1,7 @@
|
|
| 1 |
{
|
| 2 |
-
"train_runtime":
|
| 3 |
-
"train_samples_per_second":
|
| 4 |
-
"train_steps_per_second": 5.
|
| 5 |
"total_flos": 1138408789377024.0,
|
| 6 |
"train_loss": 3.6834596395492554,
|
| 7 |
"epoch": 8.0
|
|
|
|
| 1 |
{
|
| 2 |
+
"train_runtime": 71.6491,
|
| 3 |
+
"train_samples_per_second": 681.767,
|
| 4 |
+
"train_steps_per_second": 5.359,
|
| 5 |
"total_flos": 1138408789377024.0,
|
| 6 |
"train_loss": 3.6834596395492554,
|
| 7 |
"epoch": 8.0
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5969
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:de315ca70cb9dd86bda39bb42e24958d0b1fa1a5d349f0888fa94ac322554246
|
| 3 |
size 5969
|
translation_config.json
CHANGED
|
@@ -13,9 +13,9 @@
|
|
| 13 |
},
|
| 14 |
{
|
| 15 |
"path": "/tmp/dataset/cc_fake.jsonl",
|
| 16 |
-
"name": "
|
| 17 |
"weight": 0.7,
|
| 18 |
-
"priority": "
|
| 19 |
}
|
| 20 |
],
|
| 21 |
"validation_split": 0.1,
|
|
@@ -95,7 +95,7 @@
|
|
| 95 |
"predictions_file": "./predictions/sw-en-predictions.jsonl"
|
| 96 |
},
|
| 97 |
"mlflow_config": {
|
| 98 |
-
"tracking_uri": "https://
|
| 99 |
"experiment_name": "swahili-english-translation",
|
| 100 |
"run_name": "sw-en-nllb-ccaligned-production",
|
| 101 |
"log_models": true,
|
|
|
|
| 13 |
},
|
| 14 |
{
|
| 15 |
"path": "/tmp/dataset/cc_fake.jsonl",
|
| 16 |
+
"name": "ccalligned",
|
| 17 |
"weight": 0.7,
|
| 18 |
+
"priority": "high"
|
| 19 |
}
|
| 20 |
],
|
| 21 |
"validation_split": 0.1,
|
|
|
|
| 95 |
"predictions_file": "./predictions/sw-en-predictions.jsonl"
|
| 96 |
},
|
| 97 |
"mlflow_config": {
|
| 98 |
+
"tracking_uri": "https://tools-earthquake-thorough-awards.trycloudflare.com",
|
| 99 |
"experiment_name": "swahili-english-translation",
|
| 100 |
"run_name": "sw-en-nllb-ccaligned-production",
|
| 101 |
"log_models": true,
|