Upload folder using huggingface_hub
Browse files- .gitattributes +4 -0
- README.md +58 -0
- chat_template.jinja +54 -0
- checkpoint-868/chat_template.jinja +54 -0
- checkpoint-868/config.json +57 -0
- checkpoint-868/generation_config.json +13 -0
- checkpoint-868/model.safetensors +3 -0
- checkpoint-868/optimizer.pt +3 -0
- checkpoint-868/rng_state.pth +3 -0
- checkpoint-868/scheduler.pt +3 -0
- checkpoint-868/tokenizer.json +3 -0
- checkpoint-868/tokenizer_config.json +16 -0
- checkpoint-868/trainer_state.json +451 -0
- checkpoint-868/training_args.bin +3 -0
- checkpoint-992/chat_template.jinja +54 -0
- checkpoint-992/config.json +57 -0
- checkpoint-992/generation_config.json +13 -0
- checkpoint-992/model.safetensors +3 -0
- checkpoint-992/optimizer.pt +3 -0
- checkpoint-992/rng_state.pth +3 -0
- checkpoint-992/scheduler.pt +3 -0
- checkpoint-992/tokenizer.json +3 -0
- checkpoint-992/tokenizer_config.json +16 -0
- checkpoint-992/trainer_state.json +512 -0
- checkpoint-992/training_args.bin +3 -0
- checkpoint-996/chat_template.jinja +54 -0
- checkpoint-996/config.json +57 -0
- checkpoint-996/generation_config.json +13 -0
- checkpoint-996/model.safetensors +3 -0
- checkpoint-996/optimizer.pt +3 -0
- checkpoint-996/rng_state.pth +3 -0
- checkpoint-996/scheduler.pt +3 -0
- checkpoint-996/tokenizer.json +3 -0
- checkpoint-996/tokenizer_config.json +16 -0
- checkpoint-996/trainer_state.json +523 -0
- checkpoint-996/training_args.bin +3 -0
- config.json +57 -0
- generation_config.json +13 -0
- model.safetensors +3 -0
- tokenizer.json +3 -0
- tokenizer_config.json +16 -0
- training_args.bin +3 -0
.gitattributes
CHANGED
|
@@ -33,3 +33,7 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
|
| 33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
| 36 |
+
checkpoint-868/tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
| 37 |
+
checkpoint-992/tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
| 38 |
+
checkpoint-996/tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
| 39 |
+
tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
README.md
ADDED
|
@@ -0,0 +1,58 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
base_model: neuphonic/neutts-air
|
| 3 |
+
library_name: transformers
|
| 4 |
+
model_name: neutts_air_urdu_finetune
|
| 5 |
+
tags:
|
| 6 |
+
- generated_from_trainer
|
| 7 |
+
- trl
|
| 8 |
+
- sft
|
| 9 |
+
licence: license
|
| 10 |
+
---
|
| 11 |
+
|
| 12 |
+
# Model Card for neutts_air_urdu_finetune
|
| 13 |
+
|
| 14 |
+
This model is a fine-tuned version of [neuphonic/neutts-air](https://huggingface.co/neuphonic/neutts-air).
|
| 15 |
+
It has been trained using [TRL](https://github.com/huggingface/trl).
|
| 16 |
+
|
| 17 |
+
## Quick start
|
| 18 |
+
|
| 19 |
+
```python
|
| 20 |
+
from transformers import pipeline
|
| 21 |
+
|
| 22 |
+
question = "If you had a time machine, but could only go to the past or the future once and never return, which would you choose and why?"
|
| 23 |
+
generator = pipeline("text-generation", model="None", device="cuda")
|
| 24 |
+
output = generator([{"role": "user", "content": question}], max_new_tokens=128, return_full_text=False)[0]
|
| 25 |
+
print(output["generated_text"])
|
| 26 |
+
```
|
| 27 |
+
|
| 28 |
+
## Training procedure
|
| 29 |
+
|
| 30 |
+
|
| 31 |
+
|
| 32 |
+
|
| 33 |
+
|
| 34 |
+
This model was trained with SFT.
|
| 35 |
+
|
| 36 |
+
### Framework versions
|
| 37 |
+
|
| 38 |
+
- TRL: 1.9.0
|
| 39 |
+
- Transformers: 5.14.1
|
| 40 |
+
- Pytorch: 2.5.1+cu121
|
| 41 |
+
- Datasets: 2.20.0
|
| 42 |
+
- Tokenizers: 0.22.2
|
| 43 |
+
|
| 44 |
+
## Citations
|
| 45 |
+
|
| 46 |
+
|
| 47 |
+
|
| 48 |
+
Cite TRL as:
|
| 49 |
+
|
| 50 |
+
```bibtex
|
| 51 |
+
@software{vonwerra2020trl,
|
| 52 |
+
title = {{TRL: Transformers Reinforcement Learning}},
|
| 53 |
+
author = {von Werra, Leandro and Belkada, Younes and Tunstall, Lewis and Beeching, Edward and Thrush, Tristan and Lambert, Nathan and Huang, Shengyi and Rasul, Kashif and Gallouédec, Quentin},
|
| 54 |
+
license = {Apache-2.0},
|
| 55 |
+
url = {https://github.com/huggingface/trl},
|
| 56 |
+
year = {2020}
|
| 57 |
+
}
|
| 58 |
+
```
|
chat_template.jinja
ADDED
|
@@ -0,0 +1,54 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{%- if tools %}
|
| 2 |
+
{{- '<|im_start|>system\n' }}
|
| 3 |
+
{%- if messages[0]['role'] == 'system' %}
|
| 4 |
+
{{- messages[0]['content'] }}
|
| 5 |
+
{%- else %}
|
| 6 |
+
{{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}
|
| 7 |
+
{%- endif %}
|
| 8 |
+
{{- "\n\n# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
|
| 9 |
+
{%- for tool in tools %}
|
| 10 |
+
{{- "\n" }}
|
| 11 |
+
{{- tool | tojson }}
|
| 12 |
+
{%- endfor %}
|
| 13 |
+
{{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
|
| 14 |
+
{%- else %}
|
| 15 |
+
{%- if messages[0]['role'] == 'system' %}
|
| 16 |
+
{{- '<|im_start|>system\n' + messages[0]['content'] + '<|im_end|>\n' }}
|
| 17 |
+
{%- else %}
|
| 18 |
+
{{- '<|im_start|>system\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\n' }}
|
| 19 |
+
{%- endif %}
|
| 20 |
+
{%- endif %}
|
| 21 |
+
{%- for message in messages %}
|
| 22 |
+
{%- if (message.role == "user") or (message.role == "system" and not loop.first) or (message.role == "assistant" and not message.tool_calls) %}
|
| 23 |
+
{{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
|
| 24 |
+
{%- elif message.role == "assistant" %}
|
| 25 |
+
{{- '<|im_start|>' + message.role }}
|
| 26 |
+
{%- if message.content %}
|
| 27 |
+
{{- '\n' + message.content }}
|
| 28 |
+
{%- endif %}
|
| 29 |
+
{%- for tool_call in message.tool_calls %}
|
| 30 |
+
{%- if tool_call.function is defined %}
|
| 31 |
+
{%- set tool_call = tool_call.function %}
|
| 32 |
+
{%- endif %}
|
| 33 |
+
{{- '\n<tool_call>\n{"name": "' }}
|
| 34 |
+
{{- tool_call.name }}
|
| 35 |
+
{{- '", "arguments": ' }}
|
| 36 |
+
{{- tool_call.arguments | tojson }}
|
| 37 |
+
{{- '}\n</tool_call>' }}
|
| 38 |
+
{%- endfor %}
|
| 39 |
+
{{- '<|im_end|>\n' }}
|
| 40 |
+
{%- elif message.role == "tool" %}
|
| 41 |
+
{%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != "tool") %}
|
| 42 |
+
{{- '<|im_start|>user' }}
|
| 43 |
+
{%- endif %}
|
| 44 |
+
{{- '\n<tool_response>\n' }}
|
| 45 |
+
{{- message.content }}
|
| 46 |
+
{{- '\n</tool_response>' }}
|
| 47 |
+
{%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
|
| 48 |
+
{{- '<|im_end|>\n' }}
|
| 49 |
+
{%- endif %}
|
| 50 |
+
{%- endif %}
|
| 51 |
+
{%- endfor %}
|
| 52 |
+
{%- if add_generation_prompt %}
|
| 53 |
+
{{- '<|im_start|>assistant\n' }}
|
| 54 |
+
{%- endif %}
|
checkpoint-868/chat_template.jinja
ADDED
|
@@ -0,0 +1,54 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{%- if tools %}
|
| 2 |
+
{{- '<|im_start|>system\n' }}
|
| 3 |
+
{%- if messages[0]['role'] == 'system' %}
|
| 4 |
+
{{- messages[0]['content'] }}
|
| 5 |
+
{%- else %}
|
| 6 |
+
{{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}
|
| 7 |
+
{%- endif %}
|
| 8 |
+
{{- "\n\n# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
|
| 9 |
+
{%- for tool in tools %}
|
| 10 |
+
{{- "\n" }}
|
| 11 |
+
{{- tool | tojson }}
|
| 12 |
+
{%- endfor %}
|
| 13 |
+
{{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
|
| 14 |
+
{%- else %}
|
| 15 |
+
{%- if messages[0]['role'] == 'system' %}
|
| 16 |
+
{{- '<|im_start|>system\n' + messages[0]['content'] + '<|im_end|>\n' }}
|
| 17 |
+
{%- else %}
|
| 18 |
+
{{- '<|im_start|>system\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\n' }}
|
| 19 |
+
{%- endif %}
|
| 20 |
+
{%- endif %}
|
| 21 |
+
{%- for message in messages %}
|
| 22 |
+
{%- if (message.role == "user") or (message.role == "system" and not loop.first) or (message.role == "assistant" and not message.tool_calls) %}
|
| 23 |
+
{{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
|
| 24 |
+
{%- elif message.role == "assistant" %}
|
| 25 |
+
{{- '<|im_start|>' + message.role }}
|
| 26 |
+
{%- if message.content %}
|
| 27 |
+
{{- '\n' + message.content }}
|
| 28 |
+
{%- endif %}
|
| 29 |
+
{%- for tool_call in message.tool_calls %}
|
| 30 |
+
{%- if tool_call.function is defined %}
|
| 31 |
+
{%- set tool_call = tool_call.function %}
|
| 32 |
+
{%- endif %}
|
| 33 |
+
{{- '\n<tool_call>\n{"name": "' }}
|
| 34 |
+
{{- tool_call.name }}
|
| 35 |
+
{{- '", "arguments": ' }}
|
| 36 |
+
{{- tool_call.arguments | tojson }}
|
| 37 |
+
{{- '}\n</tool_call>' }}
|
| 38 |
+
{%- endfor %}
|
| 39 |
+
{{- '<|im_end|>\n' }}
|
| 40 |
+
{%- elif message.role == "tool" %}
|
| 41 |
+
{%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != "tool") %}
|
| 42 |
+
{{- '<|im_start|>user' }}
|
| 43 |
+
{%- endif %}
|
| 44 |
+
{{- '\n<tool_response>\n' }}
|
| 45 |
+
{{- message.content }}
|
| 46 |
+
{{- '\n</tool_response>' }}
|
| 47 |
+
{%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
|
| 48 |
+
{{- '<|im_end|>\n' }}
|
| 49 |
+
{%- endif %}
|
| 50 |
+
{%- endif %}
|
| 51 |
+
{%- endfor %}
|
| 52 |
+
{%- if add_generation_prompt %}
|
| 53 |
+
{{- '<|im_start|>assistant\n' }}
|
| 54 |
+
{%- endif %}
|
checkpoint-868/config.json
ADDED
|
@@ -0,0 +1,57 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"Qwen2ForCausalLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_dropout": 0.0,
|
| 6 |
+
"bos_token_id": null,
|
| 7 |
+
"dtype": "bfloat16",
|
| 8 |
+
"eos_token_id": 151645,
|
| 9 |
+
"hidden_act": "silu",
|
| 10 |
+
"hidden_size": 896,
|
| 11 |
+
"initializer_range": 0.02,
|
| 12 |
+
"intermediate_size": 4864,
|
| 13 |
+
"layer_types": [
|
| 14 |
+
"full_attention",
|
| 15 |
+
"full_attention",
|
| 16 |
+
"full_attention",
|
| 17 |
+
"full_attention",
|
| 18 |
+
"full_attention",
|
| 19 |
+
"full_attention",
|
| 20 |
+
"full_attention",
|
| 21 |
+
"full_attention",
|
| 22 |
+
"full_attention",
|
| 23 |
+
"full_attention",
|
| 24 |
+
"full_attention",
|
| 25 |
+
"full_attention",
|
| 26 |
+
"full_attention",
|
| 27 |
+
"full_attention",
|
| 28 |
+
"full_attention",
|
| 29 |
+
"full_attention",
|
| 30 |
+
"full_attention",
|
| 31 |
+
"full_attention",
|
| 32 |
+
"full_attention",
|
| 33 |
+
"full_attention",
|
| 34 |
+
"full_attention",
|
| 35 |
+
"full_attention",
|
| 36 |
+
"full_attention",
|
| 37 |
+
"full_attention"
|
| 38 |
+
],
|
| 39 |
+
"max_position_embeddings": 32768,
|
| 40 |
+
"max_window_layers": 21,
|
| 41 |
+
"model_type": "qwen2",
|
| 42 |
+
"num_attention_heads": 14,
|
| 43 |
+
"num_hidden_layers": 24,
|
| 44 |
+
"num_key_value_heads": 2,
|
| 45 |
+
"pad_token_id": 151645,
|
| 46 |
+
"rms_norm_eps": 1e-06,
|
| 47 |
+
"rope_parameters": {
|
| 48 |
+
"rope_theta": 1000000.0,
|
| 49 |
+
"rope_type": "default"
|
| 50 |
+
},
|
| 51 |
+
"sliding_window": null,
|
| 52 |
+
"tie_word_embeddings": true,
|
| 53 |
+
"transformers_version": "5.14.1",
|
| 54 |
+
"use_cache": false,
|
| 55 |
+
"use_sliding_window": false,
|
| 56 |
+
"vocab_size": 217653
|
| 57 |
+
}
|
checkpoint-868/generation_config.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"do_sample": true,
|
| 3 |
+
"eos_token_id": [
|
| 4 |
+
151645,
|
| 5 |
+
151643
|
| 6 |
+
],
|
| 7 |
+
"pad_token_id": 151645,
|
| 8 |
+
"repetition_penalty": 1.1,
|
| 9 |
+
"temperature": 0.7,
|
| 10 |
+
"top_k": 20,
|
| 11 |
+
"top_p": 0.8,
|
| 12 |
+
"transformers_version": "5.14.1"
|
| 13 |
+
}
|
checkpoint-868/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:72987f518ab398cf02ffa544b2cda2cbe222546c15f29d63c05ddd8ffc3b6103
|
| 3 |
+
size 1105862784
|
checkpoint-868/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8a4312747c7bbac8faf7745c48dee104d12b3c8ca20eb114e24b2be24e0445fd
|
| 3 |
+
size 2211903930
|
checkpoint-868/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5edb34d031c0c2b447f3eaadb401a4c1e7e7e6d8c096e28b7092e01a8bd48c92
|
| 3 |
+
size 14244
|
checkpoint-868/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1320a1bf51bd881c32f1eb4368f15fe02d9ce12d687b72e0b16c30523ed4a53d
|
| 3 |
+
size 1064
|
checkpoint-868/tokenizer.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ad19af06e958a8bdd6b45bfbba8a683aa333fc56623bd35f3a3c2b1d9c2a103e
|
| 3 |
+
size 24140418
|
checkpoint-868/tokenizer_config.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": false,
|
| 3 |
+
"backend": "tokenizers",
|
| 4 |
+
"bos_token": null,
|
| 5 |
+
"clean_up_tokenization_spaces": false,
|
| 6 |
+
"eos_token": "<|im_end|>",
|
| 7 |
+
"errors": "replace",
|
| 8 |
+
"is_local": false,
|
| 9 |
+
"local_files_only": false,
|
| 10 |
+
"model_max_length": 2048,
|
| 11 |
+
"pad_token": "<|im_end|>",
|
| 12 |
+
"padding_side": "right",
|
| 13 |
+
"split_special_tokens": false,
|
| 14 |
+
"tokenizer_class": "Qwen2Tokenizer",
|
| 15 |
+
"unk_token": null
|
| 16 |
+
}
|
checkpoint-868/trainer_state.json
ADDED
|
@@ -0,0 +1,451 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 868,
|
| 3 |
+
"best_metric": 7.3567795753479,
|
| 4 |
+
"best_model_checkpoint": "/workspace/neutts_air_urdu_finetune/checkpoint-868",
|
| 5 |
+
"epoch": 3.6779661016949152,
|
| 6 |
+
"eval_steps": 124,
|
| 7 |
+
"global_step": 868,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"entropy": 7.697032804489136,
|
| 14 |
+
"epoch": 0.1059322033898305,
|
| 15 |
+
"grad_norm": 4.78125,
|
| 16 |
+
"learning_rate": 1.6000000000000003e-05,
|
| 17 |
+
"loss": 7.989871215820313,
|
| 18 |
+
"mean_token_accuracy": 0.016311284080147745,
|
| 19 |
+
"num_tokens": 218322.0,
|
| 20 |
+
"step": 25
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"entropy": 7.771375341415405,
|
| 24 |
+
"epoch": 0.211864406779661,
|
| 25 |
+
"grad_norm": 2.671875,
|
| 26 |
+
"learning_rate": 1.9980915336317713e-05,
|
| 27 |
+
"loss": 7.803304443359375,
|
| 28 |
+
"mean_token_accuracy": 0.018622982166707516,
|
| 29 |
+
"num_tokens": 437349.0,
|
| 30 |
+
"step": 50
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"entropy": 7.646108131408692,
|
| 34 |
+
"epoch": 0.3177966101694915,
|
| 35 |
+
"grad_norm": 2.421875,
|
| 36 |
+
"learning_rate": 1.989779323028575e-05,
|
| 37 |
+
"loss": 7.629185180664063,
|
| 38 |
+
"mean_token_accuracy": 0.021061680242419244,
|
| 39 |
+
"num_tokens": 658790.0,
|
| 40 |
+
"step": 75
|
| 41 |
+
},
|
| 42 |
+
{
|
| 43 |
+
"entropy": 7.555726566314697,
|
| 44 |
+
"epoch": 0.423728813559322,
|
| 45 |
+
"grad_norm": 2.359375,
|
| 46 |
+
"learning_rate": 1.9749279121818235e-05,
|
| 47 |
+
"loss": 7.55064208984375,
|
| 48 |
+
"mean_token_accuracy": 0.022715183552354575,
|
| 49 |
+
"num_tokens": 872825.0,
|
| 50 |
+
"step": 100
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.5254237288135594,
|
| 54 |
+
"eval_entropy": 7.486619492371877,
|
| 55 |
+
"eval_loss": 7.498717784881592,
|
| 56 |
+
"eval_mean_token_accuracy": 0.02434113745888074,
|
| 57 |
+
"eval_num_tokens": 1084707.0,
|
| 58 |
+
"eval_runtime": 7.3619,
|
| 59 |
+
"eval_samples_per_second": 54.062,
|
| 60 |
+
"eval_steps_per_second": 3.396,
|
| 61 |
+
"step": 124
|
| 62 |
+
},
|
| 63 |
+
{
|
| 64 |
+
"entropy": 7.509912757873535,
|
| 65 |
+
"epoch": 0.5296610169491526,
|
| 66 |
+
"grad_norm": 2.21875,
|
| 67 |
+
"learning_rate": 1.9536354202855306e-05,
|
| 68 |
+
"loss": 7.5065460205078125,
|
| 69 |
+
"mean_token_accuracy": 0.022764644995331765,
|
| 70 |
+
"num_tokens": 1093864.0,
|
| 71 |
+
"step": 125
|
| 72 |
+
},
|
| 73 |
+
{
|
| 74 |
+
"entropy": 7.504309902191162,
|
| 75 |
+
"epoch": 0.635593220338983,
|
| 76 |
+
"grad_norm": 2.078125,
|
| 77 |
+
"learning_rate": 1.9260425209878052e-05,
|
| 78 |
+
"loss": 7.493028564453125,
|
| 79 |
+
"mean_token_accuracy": 0.022260171882808207,
|
| 80 |
+
"num_tokens": 1315198.0,
|
| 81 |
+
"step": 150
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"entropy": 7.464480457305908,
|
| 85 |
+
"epoch": 0.7415254237288136,
|
| 86 |
+
"grad_norm": 2.15625,
|
| 87 |
+
"learning_rate": 1.8923315129986838e-05,
|
| 88 |
+
"loss": 7.463182373046875,
|
| 89 |
+
"mean_token_accuracy": 0.023391987532377242,
|
| 90 |
+
"num_tokens": 1533213.0,
|
| 91 |
+
"step": 175
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"entropy": 7.452612771987915,
|
| 95 |
+
"epoch": 0.847457627118644,
|
| 96 |
+
"grad_norm": 2.34375,
|
| 97 |
+
"learning_rate": 1.8527251156925997e-05,
|
| 98 |
+
"loss": 7.447894897460937,
|
| 99 |
+
"mean_token_accuracy": 0.02403868570923805,
|
| 100 |
+
"num_tokens": 1752307.0,
|
| 101 |
+
"step": 200
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"entropy": 7.409109563827514,
|
| 105 |
+
"epoch": 0.9533898305084746,
|
| 106 |
+
"grad_norm": 1.96875,
|
| 107 |
+
"learning_rate": 1.8074849976626273e-05,
|
| 108 |
+
"loss": 7.407470703125,
|
| 109 |
+
"mean_token_accuracy": 0.023314249143004416,
|
| 110 |
+
"num_tokens": 1973350.0,
|
| 111 |
+
"step": 225
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 1.0508474576271187,
|
| 115 |
+
"eval_entropy": 7.401865323384603,
|
| 116 |
+
"eval_loss": 7.414607524871826,
|
| 117 |
+
"eval_mean_token_accuracy": 0.025272298449029524,
|
| 118 |
+
"eval_num_tokens": 2178943.0,
|
| 119 |
+
"eval_runtime": 7.4146,
|
| 120 |
+
"eval_samples_per_second": 53.678,
|
| 121 |
+
"eval_steps_per_second": 3.372,
|
| 122 |
+
"step": 248
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"entropy": 7.410350885391235,
|
| 126 |
+
"epoch": 1.0593220338983051,
|
| 127 |
+
"grad_norm": 2.078125,
|
| 128 |
+
"learning_rate": 1.756910047947926e-05,
|
| 129 |
+
"loss": 7.405702514648437,
|
| 130 |
+
"mean_token_accuracy": 0.024052796624600886,
|
| 131 |
+
"num_tokens": 2195473.0,
|
| 132 |
+
"step": 250
|
| 133 |
+
},
|
| 134 |
+
{
|
| 135 |
+
"entropy": 7.378520240783692,
|
| 136 |
+
"epoch": 1.1652542372881356,
|
| 137 |
+
"grad_norm": 2.0625,
|
| 138 |
+
"learning_rate": 1.7013344013559197e-05,
|
| 139 |
+
"loss": 7.368658447265625,
|
| 140 |
+
"mean_token_accuracy": 0.024010552018880846,
|
| 141 |
+
"num_tokens": 2410539.0,
|
| 142 |
+
"step": 275
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"entropy": 7.370533514022827,
|
| 146 |
+
"epoch": 1.271186440677966,
|
| 147 |
+
"grad_norm": 2.015625,
|
| 148 |
+
"learning_rate": 1.64112523092536e-05,
|
| 149 |
+
"loss": 7.370477905273438,
|
| 150 |
+
"mean_token_accuracy": 0.02511810462921858,
|
| 151 |
+
"num_tokens": 2630347.0,
|
| 152 |
+
"step": 300
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"entropy": 7.372100067138672,
|
| 156 |
+
"epoch": 1.3771186440677967,
|
| 157 |
+
"grad_norm": 2.140625,
|
| 158 |
+
"learning_rate": 1.5766803221148676e-05,
|
| 159 |
+
"loss": 7.374469604492187,
|
| 160 |
+
"mean_token_accuracy": 0.02450455617159605,
|
| 161 |
+
"num_tokens": 2848534.0,
|
| 162 |
+
"step": 325
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"entropy": 7.342970514297486,
|
| 166 |
+
"epoch": 1.4830508474576272,
|
| 167 |
+
"grad_norm": 1.9609375,
|
| 168 |
+
"learning_rate": 1.5084254447436169e-05,
|
| 169 |
+
"loss": 7.339466552734375,
|
| 170 |
+
"mean_token_accuracy": 0.025138991102576256,
|
| 171 |
+
"num_tokens": 3070123.0,
|
| 172 |
+
"step": 350
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 1.576271186440678,
|
| 176 |
+
"eval_entropy": 7.361576656500499,
|
| 177 |
+
"eval_loss": 7.381193161010742,
|
| 178 |
+
"eval_mean_token_accuracy": 0.025545974417279165,
|
| 179 |
+
"eval_num_tokens": 3263577.0,
|
| 180 |
+
"eval_runtime": 7.4123,
|
| 181 |
+
"eval_samples_per_second": 53.695,
|
| 182 |
+
"eval_steps_per_second": 3.373,
|
| 183 |
+
"step": 372
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"entropy": 7.361075611114502,
|
| 187 |
+
"epoch": 1.5889830508474576,
|
| 188 |
+
"grad_norm": 1.9140625,
|
| 189 |
+
"learning_rate": 1.4368115400470393e-05,
|
| 190 |
+
"loss": 7.3520880126953125,
|
| 191 |
+
"mean_token_accuracy": 0.024180141538381578,
|
| 192 |
+
"num_tokens": 3288806.0,
|
| 193 |
+
"step": 375
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"entropy": 7.3546325302124025,
|
| 197 |
+
"epoch": 1.694915254237288,
|
| 198 |
+
"grad_norm": 2.15625,
|
| 199 |
+
"learning_rate": 1.3623117414318827e-05,
|
| 200 |
+
"loss": 7.346293334960937,
|
| 201 |
+
"mean_token_accuracy": 0.024999124109745027,
|
| 202 |
+
"num_tokens": 3510776.0,
|
| 203 |
+
"step": 400
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"entropy": 7.363360710144043,
|
| 207 |
+
"epoch": 1.8008474576271185,
|
| 208 |
+
"grad_norm": 2.078125,
|
| 209 |
+
"learning_rate": 1.2854182486136944e-05,
|
| 210 |
+
"loss": 7.364286499023438,
|
| 211 |
+
"mean_token_accuracy": 0.02418241061270237,
|
| 212 |
+
"num_tokens": 3731405.0,
|
| 213 |
+
"step": 425
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"entropy": 7.334566602706909,
|
| 217 |
+
"epoch": 1.9067796610169492,
|
| 218 |
+
"grad_norm": 1.9609375,
|
| 219 |
+
"learning_rate": 1.2066390757884328e-05,
|
| 220 |
+
"loss": 7.33093505859375,
|
| 221 |
+
"mean_token_accuracy": 0.02495731335133314,
|
| 222 |
+
"num_tokens": 3953207.0,
|
| 223 |
+
"step": 450
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"entropy": 7.353079776763916,
|
| 227 |
+
"epoch": 2.01271186440678,
|
| 228 |
+
"grad_norm": 1.8671875,
|
| 229 |
+
"learning_rate": 1.1264946953221496e-05,
|
| 230 |
+
"loss": 7.351827392578125,
|
| 231 |
+
"mean_token_accuracy": 0.02447190221399069,
|
| 232 |
+
"num_tokens": 4169521.0,
|
| 233 |
+
"step": 475
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 2.1016949152542375,
|
| 237 |
+
"eval_entropy": 7.34862866004308,
|
| 238 |
+
"eval_loss": 7.365415096282959,
|
| 239 |
+
"eval_mean_token_accuracy": 0.025626841001212597,
|
| 240 |
+
"eval_num_tokens": 4351452.0,
|
| 241 |
+
"eval_runtime": 7.4142,
|
| 242 |
+
"eval_samples_per_second": 53.681,
|
| 243 |
+
"eval_steps_per_second": 3.372,
|
| 244 |
+
"step": 496
|
| 245 |
+
},
|
| 246 |
+
{
|
| 247 |
+
"entropy": 7.339797382354736,
|
| 248 |
+
"epoch": 2.1186440677966103,
|
| 249 |
+
"grad_norm": 2.03125,
|
| 250 |
+
"learning_rate": 1.0455145991329639e-05,
|
| 251 |
+
"loss": 7.3259820556640625,
|
| 252 |
+
"mean_token_accuracy": 0.025096801929175853,
|
| 253 |
+
"num_tokens": 4386912.0,
|
| 254 |
+
"step": 500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"entropy": 7.330105533599854,
|
| 258 |
+
"epoch": 2.2245762711864407,
|
| 259 |
+
"grad_norm": 2.03125,
|
| 260 |
+
"learning_rate": 9.642338004833295e-06,
|
| 261 |
+
"loss": 7.330958251953125,
|
| 262 |
+
"mean_token_accuracy": 0.024473249688744547,
|
| 263 |
+
"num_tokens": 4607819.0,
|
| 264 |
+
"step": 525
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"entropy": 7.331509189605713,
|
| 268 |
+
"epoch": 2.330508474576271,
|
| 269 |
+
"grad_norm": 1.96875,
|
| 270 |
+
"learning_rate": 8.831892992943e-06,
|
| 271 |
+
"loss": 7.329585571289062,
|
| 272 |
+
"mean_token_accuracy": 0.024803164564073086,
|
| 273 |
+
"num_tokens": 4829926.0,
|
| 274 |
+
"step": 550
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"entropy": 7.332346143722535,
|
| 278 |
+
"epoch": 2.4364406779661016,
|
| 279 |
+
"grad_norm": 1.953125,
|
| 280 |
+
"learning_rate": 8.029165343344805e-06,
|
| 281 |
+
"loss": 7.3258428955078125,
|
| 282 |
+
"mean_token_accuracy": 0.02469826426357031,
|
| 283 |
+
"num_tokens": 5045115.0,
|
| 284 |
+
"step": 575
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"entropy": 7.318505411148071,
|
| 288 |
+
"epoch": 2.542372881355932,
|
| 289 |
+
"grad_norm": 1.9140625,
|
| 290 |
+
"learning_rate": 7.2394584572309125e-06,
|
| 291 |
+
"loss": 7.3118963623046875,
|
| 292 |
+
"mean_token_accuracy": 0.02518722042441368,
|
| 293 |
+
"num_tokens": 5265448.0,
|
| 294 |
+
"step": 600
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 2.6271186440677967,
|
| 298 |
+
"eval_entropy": 7.333844860394795,
|
| 299 |
+
"eval_loss": 7.359093189239502,
|
| 300 |
+
"eval_mean_token_accuracy": 0.025550531456246972,
|
| 301 |
+
"eval_num_tokens": 5439386.0,
|
| 302 |
+
"eval_runtime": 7.4052,
|
| 303 |
+
"eval_samples_per_second": 53.746,
|
| 304 |
+
"eval_steps_per_second": 3.376,
|
| 305 |
+
"step": 620
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"entropy": 7.321087207794189,
|
| 309 |
+
"epoch": 2.648305084745763,
|
| 310 |
+
"grad_norm": 2.0625,
|
| 311 |
+
"learning_rate": 6.467989711184021e-06,
|
| 312 |
+
"loss": 7.31301513671875,
|
| 313 |
+
"mean_token_accuracy": 0.025983325839042663,
|
| 314 |
+
"num_tokens": 5482207.0,
|
| 315 |
+
"step": 625
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"entropy": 7.322469959259033,
|
| 319 |
+
"epoch": 2.7542372881355934,
|
| 320 |
+
"grad_norm": 2.125,
|
| 321 |
+
"learning_rate": 5.7198559874026945e-06,
|
| 322 |
+
"loss": 7.318033447265625,
|
| 323 |
+
"mean_token_accuracy": 0.02565582599490881,
|
| 324 |
+
"num_tokens": 5703727.0,
|
| 325 |
+
"step": 650
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"entropy": 7.31039176940918,
|
| 329 |
+
"epoch": 2.860169491525424,
|
| 330 |
+
"grad_norm": 2.03125,
|
| 331 |
+
"learning_rate": 5.000000000000003e-06,
|
| 332 |
+
"loss": 7.305034790039063,
|
| 333 |
+
"mean_token_accuracy": 0.02535436548292637,
|
| 334 |
+
"num_tokens": 5923157.0,
|
| 335 |
+
"step": 675
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"entropy": 7.3113681411743165,
|
| 339 |
+
"epoch": 2.9661016949152543,
|
| 340 |
+
"grad_norm": 1.9765625,
|
| 341 |
+
"learning_rate": 4.313177639848408e-06,
|
| 342 |
+
"loss": 7.3114166259765625,
|
| 343 |
+
"mean_token_accuracy": 0.025559407025575638,
|
| 344 |
+
"num_tokens": 6143779.0,
|
| 345 |
+
"step": 700
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"entropy": 7.3387535953521725,
|
| 349 |
+
"epoch": 3.0720338983050848,
|
| 350 |
+
"grad_norm": 1.921875,
|
| 351 |
+
"learning_rate": 3.6639265537145187e-06,
|
| 352 |
+
"loss": 7.343036499023437,
|
| 353 |
+
"mean_token_accuracy": 0.024094666354358196,
|
| 354 |
+
"num_tokens": 6365371.0,
|
| 355 |
+
"step": 725
|
| 356 |
+
},
|
| 357 |
+
{
|
| 358 |
+
"epoch": 3.152542372881356,
|
| 359 |
+
"eval_entropy": 7.332533776760101,
|
| 360 |
+
"eval_loss": 7.357120990753174,
|
| 361 |
+
"eval_mean_token_accuracy": 0.025878646954273183,
|
| 362 |
+
"eval_num_tokens": 6532337.0,
|
| 363 |
+
"eval_runtime": 7.4033,
|
| 364 |
+
"eval_samples_per_second": 53.76,
|
| 365 |
+
"eval_steps_per_second": 3.377,
|
| 366 |
+
"step": 744
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"entropy": 7.320968351364136,
|
| 370 |
+
"epoch": 3.1779661016949152,
|
| 371 |
+
"grad_norm": 2.109375,
|
| 372 |
+
"learning_rate": 3.056536165272662e-06,
|
| 373 |
+
"loss": 7.313834228515625,
|
| 374 |
+
"mean_token_accuracy": 0.025647504702210427,
|
| 375 |
+
"num_tokens": 6584143.0,
|
| 376 |
+
"step": 750
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"entropy": 7.319151067733765,
|
| 380 |
+
"epoch": 3.2838983050847457,
|
| 381 |
+
"grad_norm": 1.9140625,
|
| 382 |
+
"learning_rate": 2.4950193360603868e-06,
|
| 383 |
+
"loss": 7.315574951171875,
|
| 384 |
+
"mean_token_accuracy": 0.02516275253146887,
|
| 385 |
+
"num_tokens": 6806162.0,
|
| 386 |
+
"step": 775
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"entropy": 7.312239484786987,
|
| 390 |
+
"epoch": 3.389830508474576,
|
| 391 |
+
"grad_norm": 1.8359375,
|
| 392 |
+
"learning_rate": 1.9830858536039742e-06,
|
| 393 |
+
"loss": 7.299415893554688,
|
| 394 |
+
"mean_token_accuracy": 0.025689680464565753,
|
| 395 |
+
"num_tokens": 7025188.0,
|
| 396 |
+
"step": 800
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"entropy": 7.3349973487854,
|
| 400 |
+
"epoch": 3.4957627118644066,
|
| 401 |
+
"grad_norm": 1.90625,
|
| 402 |
+
"learning_rate": 1.524117921870789e-06,
|
| 403 |
+
"loss": 7.336126098632812,
|
| 404 |
+
"mean_token_accuracy": 0.02420500263571739,
|
| 405 |
+
"num_tokens": 7242603.0,
|
| 406 |
+
"step": 825
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"entropy": 7.306003942489624,
|
| 410 |
+
"epoch": 3.601694915254237,
|
| 411 |
+
"grad_norm": 1.765625,
|
| 412 |
+
"learning_rate": 1.121147815976248e-06,
|
| 413 |
+
"loss": 7.2987548828125,
|
| 414 |
+
"mean_token_accuracy": 0.02604618974030018,
|
| 415 |
+
"num_tokens": 7462100.0,
|
| 416 |
+
"step": 850
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 3.6779661016949152,
|
| 420 |
+
"eval_entropy": 7.331447899341583,
|
| 421 |
+
"eval_loss": 7.3567795753479,
|
| 422 |
+
"eval_mean_token_accuracy": 0.02568032095829646,
|
| 423 |
+
"eval_num_tokens": 7621408.0,
|
| 424 |
+
"eval_runtime": 7.4184,
|
| 425 |
+
"eval_samples_per_second": 53.65,
|
| 426 |
+
"eval_steps_per_second": 3.37,
|
| 427 |
+
"step": 868
|
| 428 |
+
}
|
| 429 |
+
],
|
| 430 |
+
"logging_steps": 25,
|
| 431 |
+
"max_steps": 996,
|
| 432 |
+
"num_input_tokens_seen": 0,
|
| 433 |
+
"num_train_epochs": 5,
|
| 434 |
+
"save_steps": 124,
|
| 435 |
+
"stateful_callbacks": {
|
| 436 |
+
"TrainerControl": {
|
| 437 |
+
"args": {
|
| 438 |
+
"should_epoch_stop": false,
|
| 439 |
+
"should_evaluate": false,
|
| 440 |
+
"should_log": false,
|
| 441 |
+
"should_save": true,
|
| 442 |
+
"should_training_stop": false
|
| 443 |
+
},
|
| 444 |
+
"attributes": {}
|
| 445 |
+
}
|
| 446 |
+
},
|
| 447 |
+
"total_flos": 3.817335536222208e+16,
|
| 448 |
+
"train_batch_size": 16,
|
| 449 |
+
"trial_name": null,
|
| 450 |
+
"trial_params": null
|
| 451 |
+
}
|
checkpoint-868/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:51fb563566aef1daee880df75431bb65187482b5a42194bffca0aab31b9931d2
|
| 3 |
+
size 5368
|
checkpoint-992/chat_template.jinja
ADDED
|
@@ -0,0 +1,54 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{%- if tools %}
|
| 2 |
+
{{- '<|im_start|>system\n' }}
|
| 3 |
+
{%- if messages[0]['role'] == 'system' %}
|
| 4 |
+
{{- messages[0]['content'] }}
|
| 5 |
+
{%- else %}
|
| 6 |
+
{{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}
|
| 7 |
+
{%- endif %}
|
| 8 |
+
{{- "\n\n# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
|
| 9 |
+
{%- for tool in tools %}
|
| 10 |
+
{{- "\n" }}
|
| 11 |
+
{{- tool | tojson }}
|
| 12 |
+
{%- endfor %}
|
| 13 |
+
{{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
|
| 14 |
+
{%- else %}
|
| 15 |
+
{%- if messages[0]['role'] == 'system' %}
|
| 16 |
+
{{- '<|im_start|>system\n' + messages[0]['content'] + '<|im_end|>\n' }}
|
| 17 |
+
{%- else %}
|
| 18 |
+
{{- '<|im_start|>system\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\n' }}
|
| 19 |
+
{%- endif %}
|
| 20 |
+
{%- endif %}
|
| 21 |
+
{%- for message in messages %}
|
| 22 |
+
{%- if (message.role == "user") or (message.role == "system" and not loop.first) or (message.role == "assistant" and not message.tool_calls) %}
|
| 23 |
+
{{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
|
| 24 |
+
{%- elif message.role == "assistant" %}
|
| 25 |
+
{{- '<|im_start|>' + message.role }}
|
| 26 |
+
{%- if message.content %}
|
| 27 |
+
{{- '\n' + message.content }}
|
| 28 |
+
{%- endif %}
|
| 29 |
+
{%- for tool_call in message.tool_calls %}
|
| 30 |
+
{%- if tool_call.function is defined %}
|
| 31 |
+
{%- set tool_call = tool_call.function %}
|
| 32 |
+
{%- endif %}
|
| 33 |
+
{{- '\n<tool_call>\n{"name": "' }}
|
| 34 |
+
{{- tool_call.name }}
|
| 35 |
+
{{- '", "arguments": ' }}
|
| 36 |
+
{{- tool_call.arguments | tojson }}
|
| 37 |
+
{{- '}\n</tool_call>' }}
|
| 38 |
+
{%- endfor %}
|
| 39 |
+
{{- '<|im_end|>\n' }}
|
| 40 |
+
{%- elif message.role == "tool" %}
|
| 41 |
+
{%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != "tool") %}
|
| 42 |
+
{{- '<|im_start|>user' }}
|
| 43 |
+
{%- endif %}
|
| 44 |
+
{{- '\n<tool_response>\n' }}
|
| 45 |
+
{{- message.content }}
|
| 46 |
+
{{- '\n</tool_response>' }}
|
| 47 |
+
{%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
|
| 48 |
+
{{- '<|im_end|>\n' }}
|
| 49 |
+
{%- endif %}
|
| 50 |
+
{%- endif %}
|
| 51 |
+
{%- endfor %}
|
| 52 |
+
{%- if add_generation_prompt %}
|
| 53 |
+
{{- '<|im_start|>assistant\n' }}
|
| 54 |
+
{%- endif %}
|
checkpoint-992/config.json
ADDED
|
@@ -0,0 +1,57 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"Qwen2ForCausalLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_dropout": 0.0,
|
| 6 |
+
"bos_token_id": null,
|
| 7 |
+
"dtype": "bfloat16",
|
| 8 |
+
"eos_token_id": 151645,
|
| 9 |
+
"hidden_act": "silu",
|
| 10 |
+
"hidden_size": 896,
|
| 11 |
+
"initializer_range": 0.02,
|
| 12 |
+
"intermediate_size": 4864,
|
| 13 |
+
"layer_types": [
|
| 14 |
+
"full_attention",
|
| 15 |
+
"full_attention",
|
| 16 |
+
"full_attention",
|
| 17 |
+
"full_attention",
|
| 18 |
+
"full_attention",
|
| 19 |
+
"full_attention",
|
| 20 |
+
"full_attention",
|
| 21 |
+
"full_attention",
|
| 22 |
+
"full_attention",
|
| 23 |
+
"full_attention",
|
| 24 |
+
"full_attention",
|
| 25 |
+
"full_attention",
|
| 26 |
+
"full_attention",
|
| 27 |
+
"full_attention",
|
| 28 |
+
"full_attention",
|
| 29 |
+
"full_attention",
|
| 30 |
+
"full_attention",
|
| 31 |
+
"full_attention",
|
| 32 |
+
"full_attention",
|
| 33 |
+
"full_attention",
|
| 34 |
+
"full_attention",
|
| 35 |
+
"full_attention",
|
| 36 |
+
"full_attention",
|
| 37 |
+
"full_attention"
|
| 38 |
+
],
|
| 39 |
+
"max_position_embeddings": 32768,
|
| 40 |
+
"max_window_layers": 21,
|
| 41 |
+
"model_type": "qwen2",
|
| 42 |
+
"num_attention_heads": 14,
|
| 43 |
+
"num_hidden_layers": 24,
|
| 44 |
+
"num_key_value_heads": 2,
|
| 45 |
+
"pad_token_id": 151645,
|
| 46 |
+
"rms_norm_eps": 1e-06,
|
| 47 |
+
"rope_parameters": {
|
| 48 |
+
"rope_theta": 1000000.0,
|
| 49 |
+
"rope_type": "default"
|
| 50 |
+
},
|
| 51 |
+
"sliding_window": null,
|
| 52 |
+
"tie_word_embeddings": true,
|
| 53 |
+
"transformers_version": "5.14.1",
|
| 54 |
+
"use_cache": false,
|
| 55 |
+
"use_sliding_window": false,
|
| 56 |
+
"vocab_size": 217653
|
| 57 |
+
}
|
checkpoint-992/generation_config.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"do_sample": true,
|
| 3 |
+
"eos_token_id": [
|
| 4 |
+
151645,
|
| 5 |
+
151643
|
| 6 |
+
],
|
| 7 |
+
"pad_token_id": 151645,
|
| 8 |
+
"repetition_penalty": 1.1,
|
| 9 |
+
"temperature": 0.7,
|
| 10 |
+
"top_k": 20,
|
| 11 |
+
"top_p": 0.8,
|
| 12 |
+
"transformers_version": "5.14.1"
|
| 13 |
+
}
|
checkpoint-992/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6217e5cd2f1fcb6b5775d40c6af19bef77ccfc0785719c2a3fea3c27f9d046f8
|
| 3 |
+
size 1105862784
|
checkpoint-992/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a0587cf0b401c232cd599d6a5c22c709a7ab4d53cc68d6f142a05fa70b3f29c7
|
| 3 |
+
size 2211903930
|
checkpoint-992/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7682299c566684ea51cf26f0c86b6ffaa3c0bc63cbdf84674b29a2c62ac72143
|
| 3 |
+
size 14244
|
checkpoint-992/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:582bcaa1a2856315614a378240a5cabaac9c30c01b2ac43af3afa892908adea3
|
| 3 |
+
size 1064
|
checkpoint-992/tokenizer.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ad19af06e958a8bdd6b45bfbba8a683aa333fc56623bd35f3a3c2b1d9c2a103e
|
| 3 |
+
size 24140418
|
checkpoint-992/tokenizer_config.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": false,
|
| 3 |
+
"backend": "tokenizers",
|
| 4 |
+
"bos_token": null,
|
| 5 |
+
"clean_up_tokenization_spaces": false,
|
| 6 |
+
"eos_token": "<|im_end|>",
|
| 7 |
+
"errors": "replace",
|
| 8 |
+
"is_local": false,
|
| 9 |
+
"local_files_only": false,
|
| 10 |
+
"model_max_length": 2048,
|
| 11 |
+
"pad_token": "<|im_end|>",
|
| 12 |
+
"padding_side": "right",
|
| 13 |
+
"split_special_tokens": false,
|
| 14 |
+
"tokenizer_class": "Qwen2Tokenizer",
|
| 15 |
+
"unk_token": null
|
| 16 |
+
}
|
checkpoint-992/trainer_state.json
ADDED
|
@@ -0,0 +1,512 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 992,
|
| 3 |
+
"best_metric": 7.356758117675781,
|
| 4 |
+
"best_model_checkpoint": "/workspace/neutts_air_urdu_finetune/checkpoint-992",
|
| 5 |
+
"epoch": 4.203389830508475,
|
| 6 |
+
"eval_steps": 124,
|
| 7 |
+
"global_step": 992,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"entropy": 7.697032804489136,
|
| 14 |
+
"epoch": 0.1059322033898305,
|
| 15 |
+
"grad_norm": 4.78125,
|
| 16 |
+
"learning_rate": 1.6000000000000003e-05,
|
| 17 |
+
"loss": 7.989871215820313,
|
| 18 |
+
"mean_token_accuracy": 0.016311284080147745,
|
| 19 |
+
"num_tokens": 218322.0,
|
| 20 |
+
"step": 25
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"entropy": 7.771375341415405,
|
| 24 |
+
"epoch": 0.211864406779661,
|
| 25 |
+
"grad_norm": 2.671875,
|
| 26 |
+
"learning_rate": 1.9980915336317713e-05,
|
| 27 |
+
"loss": 7.803304443359375,
|
| 28 |
+
"mean_token_accuracy": 0.018622982166707516,
|
| 29 |
+
"num_tokens": 437349.0,
|
| 30 |
+
"step": 50
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"entropy": 7.646108131408692,
|
| 34 |
+
"epoch": 0.3177966101694915,
|
| 35 |
+
"grad_norm": 2.421875,
|
| 36 |
+
"learning_rate": 1.989779323028575e-05,
|
| 37 |
+
"loss": 7.629185180664063,
|
| 38 |
+
"mean_token_accuracy": 0.021061680242419244,
|
| 39 |
+
"num_tokens": 658790.0,
|
| 40 |
+
"step": 75
|
| 41 |
+
},
|
| 42 |
+
{
|
| 43 |
+
"entropy": 7.555726566314697,
|
| 44 |
+
"epoch": 0.423728813559322,
|
| 45 |
+
"grad_norm": 2.359375,
|
| 46 |
+
"learning_rate": 1.9749279121818235e-05,
|
| 47 |
+
"loss": 7.55064208984375,
|
| 48 |
+
"mean_token_accuracy": 0.022715183552354575,
|
| 49 |
+
"num_tokens": 872825.0,
|
| 50 |
+
"step": 100
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.5254237288135594,
|
| 54 |
+
"eval_entropy": 7.486619492371877,
|
| 55 |
+
"eval_loss": 7.498717784881592,
|
| 56 |
+
"eval_mean_token_accuracy": 0.02434113745888074,
|
| 57 |
+
"eval_num_tokens": 1084707.0,
|
| 58 |
+
"eval_runtime": 7.3619,
|
| 59 |
+
"eval_samples_per_second": 54.062,
|
| 60 |
+
"eval_steps_per_second": 3.396,
|
| 61 |
+
"step": 124
|
| 62 |
+
},
|
| 63 |
+
{
|
| 64 |
+
"entropy": 7.509912757873535,
|
| 65 |
+
"epoch": 0.5296610169491526,
|
| 66 |
+
"grad_norm": 2.21875,
|
| 67 |
+
"learning_rate": 1.9536354202855306e-05,
|
| 68 |
+
"loss": 7.5065460205078125,
|
| 69 |
+
"mean_token_accuracy": 0.022764644995331765,
|
| 70 |
+
"num_tokens": 1093864.0,
|
| 71 |
+
"step": 125
|
| 72 |
+
},
|
| 73 |
+
{
|
| 74 |
+
"entropy": 7.504309902191162,
|
| 75 |
+
"epoch": 0.635593220338983,
|
| 76 |
+
"grad_norm": 2.078125,
|
| 77 |
+
"learning_rate": 1.9260425209878052e-05,
|
| 78 |
+
"loss": 7.493028564453125,
|
| 79 |
+
"mean_token_accuracy": 0.022260171882808207,
|
| 80 |
+
"num_tokens": 1315198.0,
|
| 81 |
+
"step": 150
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"entropy": 7.464480457305908,
|
| 85 |
+
"epoch": 0.7415254237288136,
|
| 86 |
+
"grad_norm": 2.15625,
|
| 87 |
+
"learning_rate": 1.8923315129986838e-05,
|
| 88 |
+
"loss": 7.463182373046875,
|
| 89 |
+
"mean_token_accuracy": 0.023391987532377242,
|
| 90 |
+
"num_tokens": 1533213.0,
|
| 91 |
+
"step": 175
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"entropy": 7.452612771987915,
|
| 95 |
+
"epoch": 0.847457627118644,
|
| 96 |
+
"grad_norm": 2.34375,
|
| 97 |
+
"learning_rate": 1.8527251156925997e-05,
|
| 98 |
+
"loss": 7.447894897460937,
|
| 99 |
+
"mean_token_accuracy": 0.02403868570923805,
|
| 100 |
+
"num_tokens": 1752307.0,
|
| 101 |
+
"step": 200
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"entropy": 7.409109563827514,
|
| 105 |
+
"epoch": 0.9533898305084746,
|
| 106 |
+
"grad_norm": 1.96875,
|
| 107 |
+
"learning_rate": 1.8074849976626273e-05,
|
| 108 |
+
"loss": 7.407470703125,
|
| 109 |
+
"mean_token_accuracy": 0.023314249143004416,
|
| 110 |
+
"num_tokens": 1973350.0,
|
| 111 |
+
"step": 225
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 1.0508474576271187,
|
| 115 |
+
"eval_entropy": 7.401865323384603,
|
| 116 |
+
"eval_loss": 7.414607524871826,
|
| 117 |
+
"eval_mean_token_accuracy": 0.025272298449029524,
|
| 118 |
+
"eval_num_tokens": 2178943.0,
|
| 119 |
+
"eval_runtime": 7.4146,
|
| 120 |
+
"eval_samples_per_second": 53.678,
|
| 121 |
+
"eval_steps_per_second": 3.372,
|
| 122 |
+
"step": 248
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"entropy": 7.410350885391235,
|
| 126 |
+
"epoch": 1.0593220338983051,
|
| 127 |
+
"grad_norm": 2.078125,
|
| 128 |
+
"learning_rate": 1.756910047947926e-05,
|
| 129 |
+
"loss": 7.405702514648437,
|
| 130 |
+
"mean_token_accuracy": 0.024052796624600886,
|
| 131 |
+
"num_tokens": 2195473.0,
|
| 132 |
+
"step": 250
|
| 133 |
+
},
|
| 134 |
+
{
|
| 135 |
+
"entropy": 7.378520240783692,
|
| 136 |
+
"epoch": 1.1652542372881356,
|
| 137 |
+
"grad_norm": 2.0625,
|
| 138 |
+
"learning_rate": 1.7013344013559197e-05,
|
| 139 |
+
"loss": 7.368658447265625,
|
| 140 |
+
"mean_token_accuracy": 0.024010552018880846,
|
| 141 |
+
"num_tokens": 2410539.0,
|
| 142 |
+
"step": 275
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"entropy": 7.370533514022827,
|
| 146 |
+
"epoch": 1.271186440677966,
|
| 147 |
+
"grad_norm": 2.015625,
|
| 148 |
+
"learning_rate": 1.64112523092536e-05,
|
| 149 |
+
"loss": 7.370477905273438,
|
| 150 |
+
"mean_token_accuracy": 0.02511810462921858,
|
| 151 |
+
"num_tokens": 2630347.0,
|
| 152 |
+
"step": 300
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"entropy": 7.372100067138672,
|
| 156 |
+
"epoch": 1.3771186440677967,
|
| 157 |
+
"grad_norm": 2.140625,
|
| 158 |
+
"learning_rate": 1.5766803221148676e-05,
|
| 159 |
+
"loss": 7.374469604492187,
|
| 160 |
+
"mean_token_accuracy": 0.02450455617159605,
|
| 161 |
+
"num_tokens": 2848534.0,
|
| 162 |
+
"step": 325
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"entropy": 7.342970514297486,
|
| 166 |
+
"epoch": 1.4830508474576272,
|
| 167 |
+
"grad_norm": 1.9609375,
|
| 168 |
+
"learning_rate": 1.5084254447436169e-05,
|
| 169 |
+
"loss": 7.339466552734375,
|
| 170 |
+
"mean_token_accuracy": 0.025138991102576256,
|
| 171 |
+
"num_tokens": 3070123.0,
|
| 172 |
+
"step": 350
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 1.576271186440678,
|
| 176 |
+
"eval_entropy": 7.361576656500499,
|
| 177 |
+
"eval_loss": 7.381193161010742,
|
| 178 |
+
"eval_mean_token_accuracy": 0.025545974417279165,
|
| 179 |
+
"eval_num_tokens": 3263577.0,
|
| 180 |
+
"eval_runtime": 7.4123,
|
| 181 |
+
"eval_samples_per_second": 53.695,
|
| 182 |
+
"eval_steps_per_second": 3.373,
|
| 183 |
+
"step": 372
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"entropy": 7.361075611114502,
|
| 187 |
+
"epoch": 1.5889830508474576,
|
| 188 |
+
"grad_norm": 1.9140625,
|
| 189 |
+
"learning_rate": 1.4368115400470393e-05,
|
| 190 |
+
"loss": 7.3520880126953125,
|
| 191 |
+
"mean_token_accuracy": 0.024180141538381578,
|
| 192 |
+
"num_tokens": 3288806.0,
|
| 193 |
+
"step": 375
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"entropy": 7.3546325302124025,
|
| 197 |
+
"epoch": 1.694915254237288,
|
| 198 |
+
"grad_norm": 2.15625,
|
| 199 |
+
"learning_rate": 1.3623117414318827e-05,
|
| 200 |
+
"loss": 7.346293334960937,
|
| 201 |
+
"mean_token_accuracy": 0.024999124109745027,
|
| 202 |
+
"num_tokens": 3510776.0,
|
| 203 |
+
"step": 400
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"entropy": 7.363360710144043,
|
| 207 |
+
"epoch": 1.8008474576271185,
|
| 208 |
+
"grad_norm": 2.078125,
|
| 209 |
+
"learning_rate": 1.2854182486136944e-05,
|
| 210 |
+
"loss": 7.364286499023438,
|
| 211 |
+
"mean_token_accuracy": 0.02418241061270237,
|
| 212 |
+
"num_tokens": 3731405.0,
|
| 213 |
+
"step": 425
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"entropy": 7.334566602706909,
|
| 217 |
+
"epoch": 1.9067796610169492,
|
| 218 |
+
"grad_norm": 1.9609375,
|
| 219 |
+
"learning_rate": 1.2066390757884328e-05,
|
| 220 |
+
"loss": 7.33093505859375,
|
| 221 |
+
"mean_token_accuracy": 0.02495731335133314,
|
| 222 |
+
"num_tokens": 3953207.0,
|
| 223 |
+
"step": 450
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"entropy": 7.353079776763916,
|
| 227 |
+
"epoch": 2.01271186440678,
|
| 228 |
+
"grad_norm": 1.8671875,
|
| 229 |
+
"learning_rate": 1.1264946953221496e-05,
|
| 230 |
+
"loss": 7.351827392578125,
|
| 231 |
+
"mean_token_accuracy": 0.02447190221399069,
|
| 232 |
+
"num_tokens": 4169521.0,
|
| 233 |
+
"step": 475
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 2.1016949152542375,
|
| 237 |
+
"eval_entropy": 7.34862866004308,
|
| 238 |
+
"eval_loss": 7.365415096282959,
|
| 239 |
+
"eval_mean_token_accuracy": 0.025626841001212597,
|
| 240 |
+
"eval_num_tokens": 4351452.0,
|
| 241 |
+
"eval_runtime": 7.4142,
|
| 242 |
+
"eval_samples_per_second": 53.681,
|
| 243 |
+
"eval_steps_per_second": 3.372,
|
| 244 |
+
"step": 496
|
| 245 |
+
},
|
| 246 |
+
{
|
| 247 |
+
"entropy": 7.339797382354736,
|
| 248 |
+
"epoch": 2.1186440677966103,
|
| 249 |
+
"grad_norm": 2.03125,
|
| 250 |
+
"learning_rate": 1.0455145991329639e-05,
|
| 251 |
+
"loss": 7.3259820556640625,
|
| 252 |
+
"mean_token_accuracy": 0.025096801929175853,
|
| 253 |
+
"num_tokens": 4386912.0,
|
| 254 |
+
"step": 500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"entropy": 7.330105533599854,
|
| 258 |
+
"epoch": 2.2245762711864407,
|
| 259 |
+
"grad_norm": 2.03125,
|
| 260 |
+
"learning_rate": 9.642338004833295e-06,
|
| 261 |
+
"loss": 7.330958251953125,
|
| 262 |
+
"mean_token_accuracy": 0.024473249688744547,
|
| 263 |
+
"num_tokens": 4607819.0,
|
| 264 |
+
"step": 525
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"entropy": 7.331509189605713,
|
| 268 |
+
"epoch": 2.330508474576271,
|
| 269 |
+
"grad_norm": 1.96875,
|
| 270 |
+
"learning_rate": 8.831892992943e-06,
|
| 271 |
+
"loss": 7.329585571289062,
|
| 272 |
+
"mean_token_accuracy": 0.024803164564073086,
|
| 273 |
+
"num_tokens": 4829926.0,
|
| 274 |
+
"step": 550
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"entropy": 7.332346143722535,
|
| 278 |
+
"epoch": 2.4364406779661016,
|
| 279 |
+
"grad_norm": 1.953125,
|
| 280 |
+
"learning_rate": 8.029165343344805e-06,
|
| 281 |
+
"loss": 7.3258428955078125,
|
| 282 |
+
"mean_token_accuracy": 0.02469826426357031,
|
| 283 |
+
"num_tokens": 5045115.0,
|
| 284 |
+
"step": 575
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"entropy": 7.318505411148071,
|
| 288 |
+
"epoch": 2.542372881355932,
|
| 289 |
+
"grad_norm": 1.9140625,
|
| 290 |
+
"learning_rate": 7.2394584572309125e-06,
|
| 291 |
+
"loss": 7.3118963623046875,
|
| 292 |
+
"mean_token_accuracy": 0.02518722042441368,
|
| 293 |
+
"num_tokens": 5265448.0,
|
| 294 |
+
"step": 600
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 2.6271186440677967,
|
| 298 |
+
"eval_entropy": 7.333844860394795,
|
| 299 |
+
"eval_loss": 7.359093189239502,
|
| 300 |
+
"eval_mean_token_accuracy": 0.025550531456246972,
|
| 301 |
+
"eval_num_tokens": 5439386.0,
|
| 302 |
+
"eval_runtime": 7.4052,
|
| 303 |
+
"eval_samples_per_second": 53.746,
|
| 304 |
+
"eval_steps_per_second": 3.376,
|
| 305 |
+
"step": 620
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"entropy": 7.321087207794189,
|
| 309 |
+
"epoch": 2.648305084745763,
|
| 310 |
+
"grad_norm": 2.0625,
|
| 311 |
+
"learning_rate": 6.467989711184021e-06,
|
| 312 |
+
"loss": 7.31301513671875,
|
| 313 |
+
"mean_token_accuracy": 0.025983325839042663,
|
| 314 |
+
"num_tokens": 5482207.0,
|
| 315 |
+
"step": 625
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"entropy": 7.322469959259033,
|
| 319 |
+
"epoch": 2.7542372881355934,
|
| 320 |
+
"grad_norm": 2.125,
|
| 321 |
+
"learning_rate": 5.7198559874026945e-06,
|
| 322 |
+
"loss": 7.318033447265625,
|
| 323 |
+
"mean_token_accuracy": 0.02565582599490881,
|
| 324 |
+
"num_tokens": 5703727.0,
|
| 325 |
+
"step": 650
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"entropy": 7.31039176940918,
|
| 329 |
+
"epoch": 2.860169491525424,
|
| 330 |
+
"grad_norm": 2.03125,
|
| 331 |
+
"learning_rate": 5.000000000000003e-06,
|
| 332 |
+
"loss": 7.305034790039063,
|
| 333 |
+
"mean_token_accuracy": 0.02535436548292637,
|
| 334 |
+
"num_tokens": 5923157.0,
|
| 335 |
+
"step": 675
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"entropy": 7.3113681411743165,
|
| 339 |
+
"epoch": 2.9661016949152543,
|
| 340 |
+
"grad_norm": 1.9765625,
|
| 341 |
+
"learning_rate": 4.313177639848408e-06,
|
| 342 |
+
"loss": 7.3114166259765625,
|
| 343 |
+
"mean_token_accuracy": 0.025559407025575638,
|
| 344 |
+
"num_tokens": 6143779.0,
|
| 345 |
+
"step": 700
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"entropy": 7.3387535953521725,
|
| 349 |
+
"epoch": 3.0720338983050848,
|
| 350 |
+
"grad_norm": 1.921875,
|
| 351 |
+
"learning_rate": 3.6639265537145187e-06,
|
| 352 |
+
"loss": 7.343036499023437,
|
| 353 |
+
"mean_token_accuracy": 0.024094666354358196,
|
| 354 |
+
"num_tokens": 6365371.0,
|
| 355 |
+
"step": 725
|
| 356 |
+
},
|
| 357 |
+
{
|
| 358 |
+
"epoch": 3.152542372881356,
|
| 359 |
+
"eval_entropy": 7.332533776760101,
|
| 360 |
+
"eval_loss": 7.357120990753174,
|
| 361 |
+
"eval_mean_token_accuracy": 0.025878646954273183,
|
| 362 |
+
"eval_num_tokens": 6532337.0,
|
| 363 |
+
"eval_runtime": 7.4033,
|
| 364 |
+
"eval_samples_per_second": 53.76,
|
| 365 |
+
"eval_steps_per_second": 3.377,
|
| 366 |
+
"step": 744
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"entropy": 7.320968351364136,
|
| 370 |
+
"epoch": 3.1779661016949152,
|
| 371 |
+
"grad_norm": 2.109375,
|
| 372 |
+
"learning_rate": 3.056536165272662e-06,
|
| 373 |
+
"loss": 7.313834228515625,
|
| 374 |
+
"mean_token_accuracy": 0.025647504702210427,
|
| 375 |
+
"num_tokens": 6584143.0,
|
| 376 |
+
"step": 750
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"entropy": 7.319151067733765,
|
| 380 |
+
"epoch": 3.2838983050847457,
|
| 381 |
+
"grad_norm": 1.9140625,
|
| 382 |
+
"learning_rate": 2.4950193360603868e-06,
|
| 383 |
+
"loss": 7.315574951171875,
|
| 384 |
+
"mean_token_accuracy": 0.02516275253146887,
|
| 385 |
+
"num_tokens": 6806162.0,
|
| 386 |
+
"step": 775
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"entropy": 7.312239484786987,
|
| 390 |
+
"epoch": 3.389830508474576,
|
| 391 |
+
"grad_norm": 1.8359375,
|
| 392 |
+
"learning_rate": 1.9830858536039742e-06,
|
| 393 |
+
"loss": 7.299415893554688,
|
| 394 |
+
"mean_token_accuracy": 0.025689680464565753,
|
| 395 |
+
"num_tokens": 7025188.0,
|
| 396 |
+
"step": 800
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"entropy": 7.3349973487854,
|
| 400 |
+
"epoch": 3.4957627118644066,
|
| 401 |
+
"grad_norm": 1.90625,
|
| 402 |
+
"learning_rate": 1.524117921870789e-06,
|
| 403 |
+
"loss": 7.336126098632812,
|
| 404 |
+
"mean_token_accuracy": 0.02420500263571739,
|
| 405 |
+
"num_tokens": 7242603.0,
|
| 406 |
+
"step": 825
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"entropy": 7.306003942489624,
|
| 410 |
+
"epoch": 3.601694915254237,
|
| 411 |
+
"grad_norm": 1.765625,
|
| 412 |
+
"learning_rate": 1.121147815976248e-06,
|
| 413 |
+
"loss": 7.2987548828125,
|
| 414 |
+
"mean_token_accuracy": 0.02604618974030018,
|
| 415 |
+
"num_tokens": 7462100.0,
|
| 416 |
+
"step": 850
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 3.6779661016949152,
|
| 420 |
+
"eval_entropy": 7.331447899341583,
|
| 421 |
+
"eval_loss": 7.3567795753479,
|
| 422 |
+
"eval_mean_token_accuracy": 0.02568032095829646,
|
| 423 |
+
"eval_num_tokens": 7621408.0,
|
| 424 |
+
"eval_runtime": 7.4184,
|
| 425 |
+
"eval_samples_per_second": 53.65,
|
| 426 |
+
"eval_steps_per_second": 3.37,
|
| 427 |
+
"step": 868
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"entropy": 7.314174528121948,
|
| 431 |
+
"epoch": 3.707627118644068,
|
| 432 |
+
"grad_norm": 2.0,
|
| 433 |
+
"learning_rate": 7.76837848774642e-07,
|
| 434 |
+
"loss": 7.314702758789062,
|
| 435 |
+
"mean_token_accuracy": 0.02541963815689087,
|
| 436 |
+
"num_tokens": 7681092.0,
|
| 437 |
+
"step": 875
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"entropy": 7.335706939697266,
|
| 441 |
+
"epoch": 3.8135593220338984,
|
| 442 |
+
"grad_norm": 2.03125,
|
| 443 |
+
"learning_rate": 4.934627816890469e-07,
|
| 444 |
+
"loss": 7.334556884765625,
|
| 445 |
+
"mean_token_accuracy": 0.02444286733865738,
|
| 446 |
+
"num_tokens": 7902109.0,
|
| 447 |
+
"step": 900
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"entropy": 7.30207275390625,
|
| 451 |
+
"epoch": 3.919491525423729,
|
| 452 |
+
"grad_norm": 1.96875,
|
| 453 |
+
"learning_rate": 2.728947959871353e-07,
|
| 454 |
+
"loss": 7.29387451171875,
|
| 455 |
+
"mean_token_accuracy": 0.02536882445216179,
|
| 456 |
+
"num_tokens": 8122890.0,
|
| 457 |
+
"step": 925
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"entropy": 7.312922801971435,
|
| 461 |
+
"epoch": 4.02542372881356,
|
| 462 |
+
"grad_norm": 1.984375,
|
| 463 |
+
"learning_rate": 1.1659112379354575e-07,
|
| 464 |
+
"loss": 7.30086181640625,
|
| 465 |
+
"mean_token_accuracy": 0.025601605214178563,
|
| 466 |
+
"num_tokens": 8341879.0,
|
| 467 |
+
"step": 950
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"entropy": 7.326199951171875,
|
| 471 |
+
"epoch": 4.13135593220339,
|
| 472 |
+
"grad_norm": 1.953125,
|
| 473 |
+
"learning_rate": 2.558442055732524e-08,
|
| 474 |
+
"loss": 7.32387939453125,
|
| 475 |
+
"mean_token_accuracy": 0.024802666790783405,
|
| 476 |
+
"num_tokens": 8564759.0,
|
| 477 |
+
"step": 975
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 4.203389830508475,
|
| 481 |
+
"eval_entropy": 7.331596851348877,
|
| 482 |
+
"eval_loss": 7.356758117675781,
|
| 483 |
+
"eval_mean_token_accuracy": 0.02577025055264433,
|
| 484 |
+
"eval_num_tokens": 8715260.0,
|
| 485 |
+
"eval_runtime": 7.397,
|
| 486 |
+
"eval_samples_per_second": 53.806,
|
| 487 |
+
"eval_steps_per_second": 3.38,
|
| 488 |
+
"step": 992
|
| 489 |
+
}
|
| 490 |
+
],
|
| 491 |
+
"logging_steps": 25,
|
| 492 |
+
"max_steps": 996,
|
| 493 |
+
"num_input_tokens_seen": 0,
|
| 494 |
+
"num_train_epochs": 5,
|
| 495 |
+
"save_steps": 124,
|
| 496 |
+
"stateful_callbacks": {
|
| 497 |
+
"TrainerControl": {
|
| 498 |
+
"args": {
|
| 499 |
+
"should_epoch_stop": false,
|
| 500 |
+
"should_evaluate": false,
|
| 501 |
+
"should_log": false,
|
| 502 |
+
"should_save": true,
|
| 503 |
+
"should_training_stop": false
|
| 504 |
+
},
|
| 505 |
+
"attributes": {}
|
| 506 |
+
}
|
| 507 |
+
},
|
| 508 |
+
"total_flos": 4.362669184253952e+16,
|
| 509 |
+
"train_batch_size": 16,
|
| 510 |
+
"trial_name": null,
|
| 511 |
+
"trial_params": null
|
| 512 |
+
}
|
checkpoint-992/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:51fb563566aef1daee880df75431bb65187482b5a42194bffca0aab31b9931d2
|
| 3 |
+
size 5368
|
checkpoint-996/chat_template.jinja
ADDED
|
@@ -0,0 +1,54 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{%- if tools %}
|
| 2 |
+
{{- '<|im_start|>system\n' }}
|
| 3 |
+
{%- if messages[0]['role'] == 'system' %}
|
| 4 |
+
{{- messages[0]['content'] }}
|
| 5 |
+
{%- else %}
|
| 6 |
+
{{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}
|
| 7 |
+
{%- endif %}
|
| 8 |
+
{{- "\n\n# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
|
| 9 |
+
{%- for tool in tools %}
|
| 10 |
+
{{- "\n" }}
|
| 11 |
+
{{- tool | tojson }}
|
| 12 |
+
{%- endfor %}
|
| 13 |
+
{{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
|
| 14 |
+
{%- else %}
|
| 15 |
+
{%- if messages[0]['role'] == 'system' %}
|
| 16 |
+
{{- '<|im_start|>system\n' + messages[0]['content'] + '<|im_end|>\n' }}
|
| 17 |
+
{%- else %}
|
| 18 |
+
{{- '<|im_start|>system\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\n' }}
|
| 19 |
+
{%- endif %}
|
| 20 |
+
{%- endif %}
|
| 21 |
+
{%- for message in messages %}
|
| 22 |
+
{%- if (message.role == "user") or (message.role == "system" and not loop.first) or (message.role == "assistant" and not message.tool_calls) %}
|
| 23 |
+
{{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
|
| 24 |
+
{%- elif message.role == "assistant" %}
|
| 25 |
+
{{- '<|im_start|>' + message.role }}
|
| 26 |
+
{%- if message.content %}
|
| 27 |
+
{{- '\n' + message.content }}
|
| 28 |
+
{%- endif %}
|
| 29 |
+
{%- for tool_call in message.tool_calls %}
|
| 30 |
+
{%- if tool_call.function is defined %}
|
| 31 |
+
{%- set tool_call = tool_call.function %}
|
| 32 |
+
{%- endif %}
|
| 33 |
+
{{- '\n<tool_call>\n{"name": "' }}
|
| 34 |
+
{{- tool_call.name }}
|
| 35 |
+
{{- '", "arguments": ' }}
|
| 36 |
+
{{- tool_call.arguments | tojson }}
|
| 37 |
+
{{- '}\n</tool_call>' }}
|
| 38 |
+
{%- endfor %}
|
| 39 |
+
{{- '<|im_end|>\n' }}
|
| 40 |
+
{%- elif message.role == "tool" %}
|
| 41 |
+
{%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != "tool") %}
|
| 42 |
+
{{- '<|im_start|>user' }}
|
| 43 |
+
{%- endif %}
|
| 44 |
+
{{- '\n<tool_response>\n' }}
|
| 45 |
+
{{- message.content }}
|
| 46 |
+
{{- '\n</tool_response>' }}
|
| 47 |
+
{%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
|
| 48 |
+
{{- '<|im_end|>\n' }}
|
| 49 |
+
{%- endif %}
|
| 50 |
+
{%- endif %}
|
| 51 |
+
{%- endfor %}
|
| 52 |
+
{%- if add_generation_prompt %}
|
| 53 |
+
{{- '<|im_start|>assistant\n' }}
|
| 54 |
+
{%- endif %}
|
checkpoint-996/config.json
ADDED
|
@@ -0,0 +1,57 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"Qwen2ForCausalLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_dropout": 0.0,
|
| 6 |
+
"bos_token_id": null,
|
| 7 |
+
"dtype": "bfloat16",
|
| 8 |
+
"eos_token_id": 151645,
|
| 9 |
+
"hidden_act": "silu",
|
| 10 |
+
"hidden_size": 896,
|
| 11 |
+
"initializer_range": 0.02,
|
| 12 |
+
"intermediate_size": 4864,
|
| 13 |
+
"layer_types": [
|
| 14 |
+
"full_attention",
|
| 15 |
+
"full_attention",
|
| 16 |
+
"full_attention",
|
| 17 |
+
"full_attention",
|
| 18 |
+
"full_attention",
|
| 19 |
+
"full_attention",
|
| 20 |
+
"full_attention",
|
| 21 |
+
"full_attention",
|
| 22 |
+
"full_attention",
|
| 23 |
+
"full_attention",
|
| 24 |
+
"full_attention",
|
| 25 |
+
"full_attention",
|
| 26 |
+
"full_attention",
|
| 27 |
+
"full_attention",
|
| 28 |
+
"full_attention",
|
| 29 |
+
"full_attention",
|
| 30 |
+
"full_attention",
|
| 31 |
+
"full_attention",
|
| 32 |
+
"full_attention",
|
| 33 |
+
"full_attention",
|
| 34 |
+
"full_attention",
|
| 35 |
+
"full_attention",
|
| 36 |
+
"full_attention",
|
| 37 |
+
"full_attention"
|
| 38 |
+
],
|
| 39 |
+
"max_position_embeddings": 32768,
|
| 40 |
+
"max_window_layers": 21,
|
| 41 |
+
"model_type": "qwen2",
|
| 42 |
+
"num_attention_heads": 14,
|
| 43 |
+
"num_hidden_layers": 24,
|
| 44 |
+
"num_key_value_heads": 2,
|
| 45 |
+
"pad_token_id": 151645,
|
| 46 |
+
"rms_norm_eps": 1e-06,
|
| 47 |
+
"rope_parameters": {
|
| 48 |
+
"rope_theta": 1000000.0,
|
| 49 |
+
"rope_type": "default"
|
| 50 |
+
},
|
| 51 |
+
"sliding_window": null,
|
| 52 |
+
"tie_word_embeddings": true,
|
| 53 |
+
"transformers_version": "5.14.1",
|
| 54 |
+
"use_cache": false,
|
| 55 |
+
"use_sliding_window": false,
|
| 56 |
+
"vocab_size": 217653
|
| 57 |
+
}
|
checkpoint-996/generation_config.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"do_sample": true,
|
| 3 |
+
"eos_token_id": [
|
| 4 |
+
151645,
|
| 5 |
+
151643
|
| 6 |
+
],
|
| 7 |
+
"pad_token_id": 151645,
|
| 8 |
+
"repetition_penalty": 1.1,
|
| 9 |
+
"temperature": 0.7,
|
| 10 |
+
"top_k": 20,
|
| 11 |
+
"top_p": 0.8,
|
| 12 |
+
"transformers_version": "5.14.1"
|
| 13 |
+
}
|
checkpoint-996/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9e6392edcbdb6cf52f7d5325d0c94ba1fe422eabad068967cb4a397130ad311f
|
| 3 |
+
size 1105862784
|
checkpoint-996/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2d7152b28c8dadbf61b30245efe964ffa3457b550ca4087b7c74b0057999e24c
|
| 3 |
+
size 2211903930
|
checkpoint-996/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8d3b7102895eb0637b0cab516bd672f216b2bf79078a83eb301011a90444f44c
|
| 3 |
+
size 14244
|
checkpoint-996/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3600ec377e6f8c3453f55e9b7986a3a096c5883da7e8b8e641766a1855e674e2
|
| 3 |
+
size 1064
|
checkpoint-996/tokenizer.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ad19af06e958a8bdd6b45bfbba8a683aa333fc56623bd35f3a3c2b1d9c2a103e
|
| 3 |
+
size 24140418
|
checkpoint-996/tokenizer_config.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": false,
|
| 3 |
+
"backend": "tokenizers",
|
| 4 |
+
"bos_token": null,
|
| 5 |
+
"clean_up_tokenization_spaces": false,
|
| 6 |
+
"eos_token": "<|im_end|>",
|
| 7 |
+
"errors": "replace",
|
| 8 |
+
"is_local": false,
|
| 9 |
+
"local_files_only": false,
|
| 10 |
+
"model_max_length": 2048,
|
| 11 |
+
"pad_token": "<|im_end|>",
|
| 12 |
+
"padding_side": "right",
|
| 13 |
+
"split_special_tokens": false,
|
| 14 |
+
"tokenizer_class": "Qwen2Tokenizer",
|
| 15 |
+
"unk_token": null
|
| 16 |
+
}
|
checkpoint-996/trainer_state.json
ADDED
|
@@ -0,0 +1,523 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 992,
|
| 3 |
+
"best_metric": 7.356758117675781,
|
| 4 |
+
"best_model_checkpoint": "/workspace/neutts_air_urdu_finetune/checkpoint-992",
|
| 5 |
+
"epoch": 4.220338983050848,
|
| 6 |
+
"eval_steps": 124,
|
| 7 |
+
"global_step": 996,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"entropy": 7.697032804489136,
|
| 14 |
+
"epoch": 0.1059322033898305,
|
| 15 |
+
"grad_norm": 4.78125,
|
| 16 |
+
"learning_rate": 1.6000000000000003e-05,
|
| 17 |
+
"loss": 7.989871215820313,
|
| 18 |
+
"mean_token_accuracy": 0.016311284080147745,
|
| 19 |
+
"num_tokens": 218322.0,
|
| 20 |
+
"step": 25
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"entropy": 7.771375341415405,
|
| 24 |
+
"epoch": 0.211864406779661,
|
| 25 |
+
"grad_norm": 2.671875,
|
| 26 |
+
"learning_rate": 1.9980915336317713e-05,
|
| 27 |
+
"loss": 7.803304443359375,
|
| 28 |
+
"mean_token_accuracy": 0.018622982166707516,
|
| 29 |
+
"num_tokens": 437349.0,
|
| 30 |
+
"step": 50
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"entropy": 7.646108131408692,
|
| 34 |
+
"epoch": 0.3177966101694915,
|
| 35 |
+
"grad_norm": 2.421875,
|
| 36 |
+
"learning_rate": 1.989779323028575e-05,
|
| 37 |
+
"loss": 7.629185180664063,
|
| 38 |
+
"mean_token_accuracy": 0.021061680242419244,
|
| 39 |
+
"num_tokens": 658790.0,
|
| 40 |
+
"step": 75
|
| 41 |
+
},
|
| 42 |
+
{
|
| 43 |
+
"entropy": 7.555726566314697,
|
| 44 |
+
"epoch": 0.423728813559322,
|
| 45 |
+
"grad_norm": 2.359375,
|
| 46 |
+
"learning_rate": 1.9749279121818235e-05,
|
| 47 |
+
"loss": 7.55064208984375,
|
| 48 |
+
"mean_token_accuracy": 0.022715183552354575,
|
| 49 |
+
"num_tokens": 872825.0,
|
| 50 |
+
"step": 100
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.5254237288135594,
|
| 54 |
+
"eval_entropy": 7.486619492371877,
|
| 55 |
+
"eval_loss": 7.498717784881592,
|
| 56 |
+
"eval_mean_token_accuracy": 0.02434113745888074,
|
| 57 |
+
"eval_num_tokens": 1084707.0,
|
| 58 |
+
"eval_runtime": 7.3619,
|
| 59 |
+
"eval_samples_per_second": 54.062,
|
| 60 |
+
"eval_steps_per_second": 3.396,
|
| 61 |
+
"step": 124
|
| 62 |
+
},
|
| 63 |
+
{
|
| 64 |
+
"entropy": 7.509912757873535,
|
| 65 |
+
"epoch": 0.5296610169491526,
|
| 66 |
+
"grad_norm": 2.21875,
|
| 67 |
+
"learning_rate": 1.9536354202855306e-05,
|
| 68 |
+
"loss": 7.5065460205078125,
|
| 69 |
+
"mean_token_accuracy": 0.022764644995331765,
|
| 70 |
+
"num_tokens": 1093864.0,
|
| 71 |
+
"step": 125
|
| 72 |
+
},
|
| 73 |
+
{
|
| 74 |
+
"entropy": 7.504309902191162,
|
| 75 |
+
"epoch": 0.635593220338983,
|
| 76 |
+
"grad_norm": 2.078125,
|
| 77 |
+
"learning_rate": 1.9260425209878052e-05,
|
| 78 |
+
"loss": 7.493028564453125,
|
| 79 |
+
"mean_token_accuracy": 0.022260171882808207,
|
| 80 |
+
"num_tokens": 1315198.0,
|
| 81 |
+
"step": 150
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"entropy": 7.464480457305908,
|
| 85 |
+
"epoch": 0.7415254237288136,
|
| 86 |
+
"grad_norm": 2.15625,
|
| 87 |
+
"learning_rate": 1.8923315129986838e-05,
|
| 88 |
+
"loss": 7.463182373046875,
|
| 89 |
+
"mean_token_accuracy": 0.023391987532377242,
|
| 90 |
+
"num_tokens": 1533213.0,
|
| 91 |
+
"step": 175
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"entropy": 7.452612771987915,
|
| 95 |
+
"epoch": 0.847457627118644,
|
| 96 |
+
"grad_norm": 2.34375,
|
| 97 |
+
"learning_rate": 1.8527251156925997e-05,
|
| 98 |
+
"loss": 7.447894897460937,
|
| 99 |
+
"mean_token_accuracy": 0.02403868570923805,
|
| 100 |
+
"num_tokens": 1752307.0,
|
| 101 |
+
"step": 200
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"entropy": 7.409109563827514,
|
| 105 |
+
"epoch": 0.9533898305084746,
|
| 106 |
+
"grad_norm": 1.96875,
|
| 107 |
+
"learning_rate": 1.8074849976626273e-05,
|
| 108 |
+
"loss": 7.407470703125,
|
| 109 |
+
"mean_token_accuracy": 0.023314249143004416,
|
| 110 |
+
"num_tokens": 1973350.0,
|
| 111 |
+
"step": 225
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 1.0508474576271187,
|
| 115 |
+
"eval_entropy": 7.401865323384603,
|
| 116 |
+
"eval_loss": 7.414607524871826,
|
| 117 |
+
"eval_mean_token_accuracy": 0.025272298449029524,
|
| 118 |
+
"eval_num_tokens": 2178943.0,
|
| 119 |
+
"eval_runtime": 7.4146,
|
| 120 |
+
"eval_samples_per_second": 53.678,
|
| 121 |
+
"eval_steps_per_second": 3.372,
|
| 122 |
+
"step": 248
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"entropy": 7.410350885391235,
|
| 126 |
+
"epoch": 1.0593220338983051,
|
| 127 |
+
"grad_norm": 2.078125,
|
| 128 |
+
"learning_rate": 1.756910047947926e-05,
|
| 129 |
+
"loss": 7.405702514648437,
|
| 130 |
+
"mean_token_accuracy": 0.024052796624600886,
|
| 131 |
+
"num_tokens": 2195473.0,
|
| 132 |
+
"step": 250
|
| 133 |
+
},
|
| 134 |
+
{
|
| 135 |
+
"entropy": 7.378520240783692,
|
| 136 |
+
"epoch": 1.1652542372881356,
|
| 137 |
+
"grad_norm": 2.0625,
|
| 138 |
+
"learning_rate": 1.7013344013559197e-05,
|
| 139 |
+
"loss": 7.368658447265625,
|
| 140 |
+
"mean_token_accuracy": 0.024010552018880846,
|
| 141 |
+
"num_tokens": 2410539.0,
|
| 142 |
+
"step": 275
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"entropy": 7.370533514022827,
|
| 146 |
+
"epoch": 1.271186440677966,
|
| 147 |
+
"grad_norm": 2.015625,
|
| 148 |
+
"learning_rate": 1.64112523092536e-05,
|
| 149 |
+
"loss": 7.370477905273438,
|
| 150 |
+
"mean_token_accuracy": 0.02511810462921858,
|
| 151 |
+
"num_tokens": 2630347.0,
|
| 152 |
+
"step": 300
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"entropy": 7.372100067138672,
|
| 156 |
+
"epoch": 1.3771186440677967,
|
| 157 |
+
"grad_norm": 2.140625,
|
| 158 |
+
"learning_rate": 1.5766803221148676e-05,
|
| 159 |
+
"loss": 7.374469604492187,
|
| 160 |
+
"mean_token_accuracy": 0.02450455617159605,
|
| 161 |
+
"num_tokens": 2848534.0,
|
| 162 |
+
"step": 325
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"entropy": 7.342970514297486,
|
| 166 |
+
"epoch": 1.4830508474576272,
|
| 167 |
+
"grad_norm": 1.9609375,
|
| 168 |
+
"learning_rate": 1.5084254447436169e-05,
|
| 169 |
+
"loss": 7.339466552734375,
|
| 170 |
+
"mean_token_accuracy": 0.025138991102576256,
|
| 171 |
+
"num_tokens": 3070123.0,
|
| 172 |
+
"step": 350
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 1.576271186440678,
|
| 176 |
+
"eval_entropy": 7.361576656500499,
|
| 177 |
+
"eval_loss": 7.381193161010742,
|
| 178 |
+
"eval_mean_token_accuracy": 0.025545974417279165,
|
| 179 |
+
"eval_num_tokens": 3263577.0,
|
| 180 |
+
"eval_runtime": 7.4123,
|
| 181 |
+
"eval_samples_per_second": 53.695,
|
| 182 |
+
"eval_steps_per_second": 3.373,
|
| 183 |
+
"step": 372
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"entropy": 7.361075611114502,
|
| 187 |
+
"epoch": 1.5889830508474576,
|
| 188 |
+
"grad_norm": 1.9140625,
|
| 189 |
+
"learning_rate": 1.4368115400470393e-05,
|
| 190 |
+
"loss": 7.3520880126953125,
|
| 191 |
+
"mean_token_accuracy": 0.024180141538381578,
|
| 192 |
+
"num_tokens": 3288806.0,
|
| 193 |
+
"step": 375
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"entropy": 7.3546325302124025,
|
| 197 |
+
"epoch": 1.694915254237288,
|
| 198 |
+
"grad_norm": 2.15625,
|
| 199 |
+
"learning_rate": 1.3623117414318827e-05,
|
| 200 |
+
"loss": 7.346293334960937,
|
| 201 |
+
"mean_token_accuracy": 0.024999124109745027,
|
| 202 |
+
"num_tokens": 3510776.0,
|
| 203 |
+
"step": 400
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"entropy": 7.363360710144043,
|
| 207 |
+
"epoch": 1.8008474576271185,
|
| 208 |
+
"grad_norm": 2.078125,
|
| 209 |
+
"learning_rate": 1.2854182486136944e-05,
|
| 210 |
+
"loss": 7.364286499023438,
|
| 211 |
+
"mean_token_accuracy": 0.02418241061270237,
|
| 212 |
+
"num_tokens": 3731405.0,
|
| 213 |
+
"step": 425
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"entropy": 7.334566602706909,
|
| 217 |
+
"epoch": 1.9067796610169492,
|
| 218 |
+
"grad_norm": 1.9609375,
|
| 219 |
+
"learning_rate": 1.2066390757884328e-05,
|
| 220 |
+
"loss": 7.33093505859375,
|
| 221 |
+
"mean_token_accuracy": 0.02495731335133314,
|
| 222 |
+
"num_tokens": 3953207.0,
|
| 223 |
+
"step": 450
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"entropy": 7.353079776763916,
|
| 227 |
+
"epoch": 2.01271186440678,
|
| 228 |
+
"grad_norm": 1.8671875,
|
| 229 |
+
"learning_rate": 1.1264946953221496e-05,
|
| 230 |
+
"loss": 7.351827392578125,
|
| 231 |
+
"mean_token_accuracy": 0.02447190221399069,
|
| 232 |
+
"num_tokens": 4169521.0,
|
| 233 |
+
"step": 475
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 2.1016949152542375,
|
| 237 |
+
"eval_entropy": 7.34862866004308,
|
| 238 |
+
"eval_loss": 7.365415096282959,
|
| 239 |
+
"eval_mean_token_accuracy": 0.025626841001212597,
|
| 240 |
+
"eval_num_tokens": 4351452.0,
|
| 241 |
+
"eval_runtime": 7.4142,
|
| 242 |
+
"eval_samples_per_second": 53.681,
|
| 243 |
+
"eval_steps_per_second": 3.372,
|
| 244 |
+
"step": 496
|
| 245 |
+
},
|
| 246 |
+
{
|
| 247 |
+
"entropy": 7.339797382354736,
|
| 248 |
+
"epoch": 2.1186440677966103,
|
| 249 |
+
"grad_norm": 2.03125,
|
| 250 |
+
"learning_rate": 1.0455145991329639e-05,
|
| 251 |
+
"loss": 7.3259820556640625,
|
| 252 |
+
"mean_token_accuracy": 0.025096801929175853,
|
| 253 |
+
"num_tokens": 4386912.0,
|
| 254 |
+
"step": 500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"entropy": 7.330105533599854,
|
| 258 |
+
"epoch": 2.2245762711864407,
|
| 259 |
+
"grad_norm": 2.03125,
|
| 260 |
+
"learning_rate": 9.642338004833295e-06,
|
| 261 |
+
"loss": 7.330958251953125,
|
| 262 |
+
"mean_token_accuracy": 0.024473249688744547,
|
| 263 |
+
"num_tokens": 4607819.0,
|
| 264 |
+
"step": 525
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"entropy": 7.331509189605713,
|
| 268 |
+
"epoch": 2.330508474576271,
|
| 269 |
+
"grad_norm": 1.96875,
|
| 270 |
+
"learning_rate": 8.831892992943e-06,
|
| 271 |
+
"loss": 7.329585571289062,
|
| 272 |
+
"mean_token_accuracy": 0.024803164564073086,
|
| 273 |
+
"num_tokens": 4829926.0,
|
| 274 |
+
"step": 550
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"entropy": 7.332346143722535,
|
| 278 |
+
"epoch": 2.4364406779661016,
|
| 279 |
+
"grad_norm": 1.953125,
|
| 280 |
+
"learning_rate": 8.029165343344805e-06,
|
| 281 |
+
"loss": 7.3258428955078125,
|
| 282 |
+
"mean_token_accuracy": 0.02469826426357031,
|
| 283 |
+
"num_tokens": 5045115.0,
|
| 284 |
+
"step": 575
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"entropy": 7.318505411148071,
|
| 288 |
+
"epoch": 2.542372881355932,
|
| 289 |
+
"grad_norm": 1.9140625,
|
| 290 |
+
"learning_rate": 7.2394584572309125e-06,
|
| 291 |
+
"loss": 7.3118963623046875,
|
| 292 |
+
"mean_token_accuracy": 0.02518722042441368,
|
| 293 |
+
"num_tokens": 5265448.0,
|
| 294 |
+
"step": 600
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 2.6271186440677967,
|
| 298 |
+
"eval_entropy": 7.333844860394795,
|
| 299 |
+
"eval_loss": 7.359093189239502,
|
| 300 |
+
"eval_mean_token_accuracy": 0.025550531456246972,
|
| 301 |
+
"eval_num_tokens": 5439386.0,
|
| 302 |
+
"eval_runtime": 7.4052,
|
| 303 |
+
"eval_samples_per_second": 53.746,
|
| 304 |
+
"eval_steps_per_second": 3.376,
|
| 305 |
+
"step": 620
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"entropy": 7.321087207794189,
|
| 309 |
+
"epoch": 2.648305084745763,
|
| 310 |
+
"grad_norm": 2.0625,
|
| 311 |
+
"learning_rate": 6.467989711184021e-06,
|
| 312 |
+
"loss": 7.31301513671875,
|
| 313 |
+
"mean_token_accuracy": 0.025983325839042663,
|
| 314 |
+
"num_tokens": 5482207.0,
|
| 315 |
+
"step": 625
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"entropy": 7.322469959259033,
|
| 319 |
+
"epoch": 2.7542372881355934,
|
| 320 |
+
"grad_norm": 2.125,
|
| 321 |
+
"learning_rate": 5.7198559874026945e-06,
|
| 322 |
+
"loss": 7.318033447265625,
|
| 323 |
+
"mean_token_accuracy": 0.02565582599490881,
|
| 324 |
+
"num_tokens": 5703727.0,
|
| 325 |
+
"step": 650
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"entropy": 7.31039176940918,
|
| 329 |
+
"epoch": 2.860169491525424,
|
| 330 |
+
"grad_norm": 2.03125,
|
| 331 |
+
"learning_rate": 5.000000000000003e-06,
|
| 332 |
+
"loss": 7.305034790039063,
|
| 333 |
+
"mean_token_accuracy": 0.02535436548292637,
|
| 334 |
+
"num_tokens": 5923157.0,
|
| 335 |
+
"step": 675
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"entropy": 7.3113681411743165,
|
| 339 |
+
"epoch": 2.9661016949152543,
|
| 340 |
+
"grad_norm": 1.9765625,
|
| 341 |
+
"learning_rate": 4.313177639848408e-06,
|
| 342 |
+
"loss": 7.3114166259765625,
|
| 343 |
+
"mean_token_accuracy": 0.025559407025575638,
|
| 344 |
+
"num_tokens": 6143779.0,
|
| 345 |
+
"step": 700
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"entropy": 7.3387535953521725,
|
| 349 |
+
"epoch": 3.0720338983050848,
|
| 350 |
+
"grad_norm": 1.921875,
|
| 351 |
+
"learning_rate": 3.6639265537145187e-06,
|
| 352 |
+
"loss": 7.343036499023437,
|
| 353 |
+
"mean_token_accuracy": 0.024094666354358196,
|
| 354 |
+
"num_tokens": 6365371.0,
|
| 355 |
+
"step": 725
|
| 356 |
+
},
|
| 357 |
+
{
|
| 358 |
+
"epoch": 3.152542372881356,
|
| 359 |
+
"eval_entropy": 7.332533776760101,
|
| 360 |
+
"eval_loss": 7.357120990753174,
|
| 361 |
+
"eval_mean_token_accuracy": 0.025878646954273183,
|
| 362 |
+
"eval_num_tokens": 6532337.0,
|
| 363 |
+
"eval_runtime": 7.4033,
|
| 364 |
+
"eval_samples_per_second": 53.76,
|
| 365 |
+
"eval_steps_per_second": 3.377,
|
| 366 |
+
"step": 744
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"entropy": 7.320968351364136,
|
| 370 |
+
"epoch": 3.1779661016949152,
|
| 371 |
+
"grad_norm": 2.109375,
|
| 372 |
+
"learning_rate": 3.056536165272662e-06,
|
| 373 |
+
"loss": 7.313834228515625,
|
| 374 |
+
"mean_token_accuracy": 0.025647504702210427,
|
| 375 |
+
"num_tokens": 6584143.0,
|
| 376 |
+
"step": 750
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"entropy": 7.319151067733765,
|
| 380 |
+
"epoch": 3.2838983050847457,
|
| 381 |
+
"grad_norm": 1.9140625,
|
| 382 |
+
"learning_rate": 2.4950193360603868e-06,
|
| 383 |
+
"loss": 7.315574951171875,
|
| 384 |
+
"mean_token_accuracy": 0.02516275253146887,
|
| 385 |
+
"num_tokens": 6806162.0,
|
| 386 |
+
"step": 775
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"entropy": 7.312239484786987,
|
| 390 |
+
"epoch": 3.389830508474576,
|
| 391 |
+
"grad_norm": 1.8359375,
|
| 392 |
+
"learning_rate": 1.9830858536039742e-06,
|
| 393 |
+
"loss": 7.299415893554688,
|
| 394 |
+
"mean_token_accuracy": 0.025689680464565753,
|
| 395 |
+
"num_tokens": 7025188.0,
|
| 396 |
+
"step": 800
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"entropy": 7.3349973487854,
|
| 400 |
+
"epoch": 3.4957627118644066,
|
| 401 |
+
"grad_norm": 1.90625,
|
| 402 |
+
"learning_rate": 1.524117921870789e-06,
|
| 403 |
+
"loss": 7.336126098632812,
|
| 404 |
+
"mean_token_accuracy": 0.02420500263571739,
|
| 405 |
+
"num_tokens": 7242603.0,
|
| 406 |
+
"step": 825
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"entropy": 7.306003942489624,
|
| 410 |
+
"epoch": 3.601694915254237,
|
| 411 |
+
"grad_norm": 1.765625,
|
| 412 |
+
"learning_rate": 1.121147815976248e-06,
|
| 413 |
+
"loss": 7.2987548828125,
|
| 414 |
+
"mean_token_accuracy": 0.02604618974030018,
|
| 415 |
+
"num_tokens": 7462100.0,
|
| 416 |
+
"step": 850
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 3.6779661016949152,
|
| 420 |
+
"eval_entropy": 7.331447899341583,
|
| 421 |
+
"eval_loss": 7.3567795753479,
|
| 422 |
+
"eval_mean_token_accuracy": 0.02568032095829646,
|
| 423 |
+
"eval_num_tokens": 7621408.0,
|
| 424 |
+
"eval_runtime": 7.4184,
|
| 425 |
+
"eval_samples_per_second": 53.65,
|
| 426 |
+
"eval_steps_per_second": 3.37,
|
| 427 |
+
"step": 868
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"entropy": 7.314174528121948,
|
| 431 |
+
"epoch": 3.707627118644068,
|
| 432 |
+
"grad_norm": 2.0,
|
| 433 |
+
"learning_rate": 7.76837848774642e-07,
|
| 434 |
+
"loss": 7.314702758789062,
|
| 435 |
+
"mean_token_accuracy": 0.02541963815689087,
|
| 436 |
+
"num_tokens": 7681092.0,
|
| 437 |
+
"step": 875
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"entropy": 7.335706939697266,
|
| 441 |
+
"epoch": 3.8135593220338984,
|
| 442 |
+
"grad_norm": 2.03125,
|
| 443 |
+
"learning_rate": 4.934627816890469e-07,
|
| 444 |
+
"loss": 7.334556884765625,
|
| 445 |
+
"mean_token_accuracy": 0.02444286733865738,
|
| 446 |
+
"num_tokens": 7902109.0,
|
| 447 |
+
"step": 900
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"entropy": 7.30207275390625,
|
| 451 |
+
"epoch": 3.919491525423729,
|
| 452 |
+
"grad_norm": 1.96875,
|
| 453 |
+
"learning_rate": 2.728947959871353e-07,
|
| 454 |
+
"loss": 7.29387451171875,
|
| 455 |
+
"mean_token_accuracy": 0.02536882445216179,
|
| 456 |
+
"num_tokens": 8122890.0,
|
| 457 |
+
"step": 925
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"entropy": 7.312922801971435,
|
| 461 |
+
"epoch": 4.02542372881356,
|
| 462 |
+
"grad_norm": 1.984375,
|
| 463 |
+
"learning_rate": 1.1659112379354575e-07,
|
| 464 |
+
"loss": 7.30086181640625,
|
| 465 |
+
"mean_token_accuracy": 0.025601605214178563,
|
| 466 |
+
"num_tokens": 8341879.0,
|
| 467 |
+
"step": 950
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"entropy": 7.326199951171875,
|
| 471 |
+
"epoch": 4.13135593220339,
|
| 472 |
+
"grad_norm": 1.953125,
|
| 473 |
+
"learning_rate": 2.558442055732524e-08,
|
| 474 |
+
"loss": 7.32387939453125,
|
| 475 |
+
"mean_token_accuracy": 0.024802666790783405,
|
| 476 |
+
"num_tokens": 8564759.0,
|
| 477 |
+
"step": 975
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 4.203389830508475,
|
| 481 |
+
"eval_entropy": 7.331596851348877,
|
| 482 |
+
"eval_loss": 7.356758117675781,
|
| 483 |
+
"eval_mean_token_accuracy": 0.02577025055264433,
|
| 484 |
+
"eval_num_tokens": 8715260.0,
|
| 485 |
+
"eval_runtime": 7.397,
|
| 486 |
+
"eval_samples_per_second": 53.806,
|
| 487 |
+
"eval_steps_per_second": 3.38,
|
| 488 |
+
"step": 992
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 4.220338983050848,
|
| 492 |
+
"eval_entropy": 7.331597983837128,
|
| 493 |
+
"eval_loss": 7.3567633628845215,
|
| 494 |
+
"eval_mean_token_accuracy": 0.02577025055264433,
|
| 495 |
+
"eval_num_tokens": 8750612.0,
|
| 496 |
+
"eval_runtime": 7.4178,
|
| 497 |
+
"eval_samples_per_second": 53.655,
|
| 498 |
+
"eval_steps_per_second": 3.37,
|
| 499 |
+
"step": 996
|
| 500 |
+
}
|
| 501 |
+
],
|
| 502 |
+
"logging_steps": 25,
|
| 503 |
+
"max_steps": 996,
|
| 504 |
+
"num_input_tokens_seen": 0,
|
| 505 |
+
"num_train_epochs": 5,
|
| 506 |
+
"save_steps": 124,
|
| 507 |
+
"stateful_callbacks": {
|
| 508 |
+
"TrainerControl": {
|
| 509 |
+
"args": {
|
| 510 |
+
"should_epoch_stop": false,
|
| 511 |
+
"should_evaluate": false,
|
| 512 |
+
"should_log": false,
|
| 513 |
+
"should_save": true,
|
| 514 |
+
"should_training_stop": true
|
| 515 |
+
},
|
| 516 |
+
"attributes": {}
|
| 517 |
+
}
|
| 518 |
+
},
|
| 519 |
+
"total_flos": 4.380260592254976e+16,
|
| 520 |
+
"train_batch_size": 16,
|
| 521 |
+
"trial_name": null,
|
| 522 |
+
"trial_params": null
|
| 523 |
+
}
|
checkpoint-996/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:51fb563566aef1daee880df75431bb65187482b5a42194bffca0aab31b9931d2
|
| 3 |
+
size 5368
|
config.json
ADDED
|
@@ -0,0 +1,57 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"Qwen2ForCausalLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_dropout": 0.0,
|
| 6 |
+
"bos_token_id": null,
|
| 7 |
+
"dtype": "bfloat16",
|
| 8 |
+
"eos_token_id": 151645,
|
| 9 |
+
"hidden_act": "silu",
|
| 10 |
+
"hidden_size": 896,
|
| 11 |
+
"initializer_range": 0.02,
|
| 12 |
+
"intermediate_size": 4864,
|
| 13 |
+
"layer_types": [
|
| 14 |
+
"full_attention",
|
| 15 |
+
"full_attention",
|
| 16 |
+
"full_attention",
|
| 17 |
+
"full_attention",
|
| 18 |
+
"full_attention",
|
| 19 |
+
"full_attention",
|
| 20 |
+
"full_attention",
|
| 21 |
+
"full_attention",
|
| 22 |
+
"full_attention",
|
| 23 |
+
"full_attention",
|
| 24 |
+
"full_attention",
|
| 25 |
+
"full_attention",
|
| 26 |
+
"full_attention",
|
| 27 |
+
"full_attention",
|
| 28 |
+
"full_attention",
|
| 29 |
+
"full_attention",
|
| 30 |
+
"full_attention",
|
| 31 |
+
"full_attention",
|
| 32 |
+
"full_attention",
|
| 33 |
+
"full_attention",
|
| 34 |
+
"full_attention",
|
| 35 |
+
"full_attention",
|
| 36 |
+
"full_attention",
|
| 37 |
+
"full_attention"
|
| 38 |
+
],
|
| 39 |
+
"max_position_embeddings": 32768,
|
| 40 |
+
"max_window_layers": 21,
|
| 41 |
+
"model_type": "qwen2",
|
| 42 |
+
"num_attention_heads": 14,
|
| 43 |
+
"num_hidden_layers": 24,
|
| 44 |
+
"num_key_value_heads": 2,
|
| 45 |
+
"pad_token_id": 151645,
|
| 46 |
+
"rms_norm_eps": 1e-06,
|
| 47 |
+
"rope_parameters": {
|
| 48 |
+
"rope_theta": 1000000.0,
|
| 49 |
+
"rope_type": "default"
|
| 50 |
+
},
|
| 51 |
+
"sliding_window": null,
|
| 52 |
+
"tie_word_embeddings": true,
|
| 53 |
+
"transformers_version": "5.14.1",
|
| 54 |
+
"use_cache": false,
|
| 55 |
+
"use_sliding_window": false,
|
| 56 |
+
"vocab_size": 217653
|
| 57 |
+
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"do_sample": true,
|
| 3 |
+
"eos_token_id": [
|
| 4 |
+
151645,
|
| 5 |
+
151643
|
| 6 |
+
],
|
| 7 |
+
"pad_token_id": 151645,
|
| 8 |
+
"repetition_penalty": 1.1,
|
| 9 |
+
"temperature": 0.7,
|
| 10 |
+
"top_k": 20,
|
| 11 |
+
"top_p": 0.8,
|
| 12 |
+
"transformers_version": "5.14.1"
|
| 13 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6217e5cd2f1fcb6b5775d40c6af19bef77ccfc0785719c2a3fea3c27f9d046f8
|
| 3 |
+
size 1105862784
|
tokenizer.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ad19af06e958a8bdd6b45bfbba8a683aa333fc56623bd35f3a3c2b1d9c2a103e
|
| 3 |
+
size 24140418
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": false,
|
| 3 |
+
"backend": "tokenizers",
|
| 4 |
+
"bos_token": null,
|
| 5 |
+
"clean_up_tokenization_spaces": false,
|
| 6 |
+
"eos_token": "<|im_end|>",
|
| 7 |
+
"errors": "replace",
|
| 8 |
+
"is_local": false,
|
| 9 |
+
"local_files_only": false,
|
| 10 |
+
"model_max_length": 2048,
|
| 11 |
+
"pad_token": "<|im_end|>",
|
| 12 |
+
"padding_side": "right",
|
| 13 |
+
"split_special_tokens": false,
|
| 14 |
+
"tokenizer_class": "Qwen2Tokenizer",
|
| 15 |
+
"unk_token": null
|
| 16 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:51fb563566aef1daee880df75431bb65187482b5a42194bffca0aab31b9931d2
|
| 3 |
+
size 5368
|