Commit ·
1fa4da1
1
Parent(s): 072dc71
Upload folder using huggingface_hub
Browse files- README.md +21 -0
- adapter_config.json +25 -0
- adapter_model.bin +3 -0
- optimizer.pt +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- special_tokens_map.json +11 -0
- tokenizer.json +0 -0
- tokenizer_config.json +44 -0
- trainer_state.json +751 -0
- training_args.bin +3 -0
README.md
ADDED
|
@@ -0,0 +1,21 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: peft
|
| 3 |
+
---
|
| 4 |
+
## Training procedure
|
| 5 |
+
|
| 6 |
+
|
| 7 |
+
The following `bitsandbytes` quantization config was used during training:
|
| 8 |
+
- quant_method: bitsandbytes
|
| 9 |
+
- load_in_8bit: False
|
| 10 |
+
- load_in_4bit: True
|
| 11 |
+
- llm_int8_threshold: 6.0
|
| 12 |
+
- llm_int8_skip_modules: None
|
| 13 |
+
- llm_int8_enable_fp32_cpu_offload: False
|
| 14 |
+
- llm_int8_has_fp16_weight: False
|
| 15 |
+
- bnb_4bit_quant_type: fp4
|
| 16 |
+
- bnb_4bit_use_double_quant: False
|
| 17 |
+
- bnb_4bit_compute_dtype: float32
|
| 18 |
+
### Framework versions
|
| 19 |
+
|
| 20 |
+
|
| 21 |
+
- PEFT 0.5.0
|
adapter_config.json
ADDED
|
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"auto_mapping": null,
|
| 3 |
+
"base_model_name_or_path": null,
|
| 4 |
+
"bias": "none",
|
| 5 |
+
"fan_in_fan_out": false,
|
| 6 |
+
"inference_mode": true,
|
| 7 |
+
"init_lora_weights": true,
|
| 8 |
+
"layers_pattern": null,
|
| 9 |
+
"layers_to_transform": null,
|
| 10 |
+
"lora_alpha": 16,
|
| 11 |
+
"lora_dropout": 0.05,
|
| 12 |
+
"modules_to_save": null,
|
| 13 |
+
"peft_type": "LORA",
|
| 14 |
+
"r": 8,
|
| 15 |
+
"revision": null,
|
| 16 |
+
"target_modules": [
|
| 17 |
+
"q_proj",
|
| 18 |
+
"k_proj",
|
| 19 |
+
"v_proj",
|
| 20 |
+
"gate_proj",
|
| 21 |
+
"up_proj",
|
| 22 |
+
"down_proj"
|
| 23 |
+
],
|
| 24 |
+
"task_type": "CAUSAL_LM"
|
| 25 |
+
}
|
adapter_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7fb41ed248b38a0d5c018c6a8583bdfba80b6d2fa606506f6813c43958a99bbf
|
| 3 |
+
size 75641741
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:446b771342b713e07f4acb65267ef0679ef125566f0de980a5ac69b427e4f799
|
| 3 |
+
size 151224453
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:972139d83957a9cf2600cb6eeca17287d7a5377c33a53500ae7e13fe830ad36b
|
| 3 |
+
size 14575
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f03c5a9d9fd80414287e17e83b0d9b80cfdcf5ad7a4a9d63da800e7a44f10384
|
| 3 |
+
size 627
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,11 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"additional_special_tokens": [
|
| 3 |
+
"<unk>",
|
| 4 |
+
"<s>",
|
| 5 |
+
"</s>"
|
| 6 |
+
],
|
| 7 |
+
"bos_token": "<s>",
|
| 8 |
+
"eos_token": "</s>",
|
| 9 |
+
"pad_token": "</s>",
|
| 10 |
+
"unk_token": "<unk>"
|
| 11 |
+
}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,44 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "<unk>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<s>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "</s>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
}
|
| 27 |
+
},
|
| 28 |
+
"additional_special_tokens": [
|
| 29 |
+
"<unk>",
|
| 30 |
+
"<s>",
|
| 31 |
+
"</s>"
|
| 32 |
+
],
|
| 33 |
+
"bos_token": "<s>",
|
| 34 |
+
"clean_up_tokenization_spaces": false,
|
| 35 |
+
"eos_token": "</s>",
|
| 36 |
+
"legacy": true,
|
| 37 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 38 |
+
"pad_token": "</s>",
|
| 39 |
+
"sp_model_kwargs": {},
|
| 40 |
+
"spaces_between_special_tokens": false,
|
| 41 |
+
"tokenizer_class": "LlamaTokenizer",
|
| 42 |
+
"unk_token": "<unk>",
|
| 43 |
+
"use_default_system_prompt": true
|
| 44 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,751 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.09665104141497124,
|
| 5 |
+
"eval_steps": 200,
|
| 6 |
+
"global_step": 500,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.0,
|
| 13 |
+
"learning_rate": 9.652509652509653e-06,
|
| 14 |
+
"logits/chosen": -3.11246657371521,
|
| 15 |
+
"logits/rejected": -3.086373805999756,
|
| 16 |
+
"logps/chosen": -113.73238372802734,
|
| 17 |
+
"logps/rejected": -109.32698822021484,
|
| 18 |
+
"loss": 0.721,
|
| 19 |
+
"rewards/accuracies": 0.44999998807907104,
|
| 20 |
+
"rewards/chosen": -0.2993558943271637,
|
| 21 |
+
"rewards/margins": -0.0277109295129776,
|
| 22 |
+
"rewards/rejected": -0.2716449797153473,
|
| 23 |
+
"step": 10
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.0,
|
| 27 |
+
"learning_rate": 1.9305019305019306e-05,
|
| 28 |
+
"logits/chosen": -3.110931873321533,
|
| 29 |
+
"logits/rejected": -3.1171531677246094,
|
| 30 |
+
"logps/chosen": -122.4663314819336,
|
| 31 |
+
"logps/rejected": -113.23054504394531,
|
| 32 |
+
"loss": 0.7161,
|
| 33 |
+
"rewards/accuracies": 0.42500001192092896,
|
| 34 |
+
"rewards/chosen": -0.2805718183517456,
|
| 35 |
+
"rewards/margins": -0.03167964145541191,
|
| 36 |
+
"rewards/rejected": -0.2488921880722046,
|
| 37 |
+
"step": 20
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.01,
|
| 41 |
+
"learning_rate": 2.895752895752896e-05,
|
| 42 |
+
"logits/chosen": -3.139052629470825,
|
| 43 |
+
"logits/rejected": -3.1156527996063232,
|
| 44 |
+
"logps/chosen": -126.01689147949219,
|
| 45 |
+
"logps/rejected": -100.77046203613281,
|
| 46 |
+
"loss": 0.7224,
|
| 47 |
+
"rewards/accuracies": 0.4625000059604645,
|
| 48 |
+
"rewards/chosen": -0.3138067126274109,
|
| 49 |
+
"rewards/margins": -0.038806475698947906,
|
| 50 |
+
"rewards/rejected": -0.2750001847743988,
|
| 51 |
+
"step": 30
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.01,
|
| 55 |
+
"learning_rate": 3.764478764478765e-05,
|
| 56 |
+
"logits/chosen": -3.155150890350342,
|
| 57 |
+
"logits/rejected": -3.1715970039367676,
|
| 58 |
+
"logps/chosen": -133.27737426757812,
|
| 59 |
+
"logps/rejected": -118.9439926147461,
|
| 60 |
+
"loss": 0.713,
|
| 61 |
+
"rewards/accuracies": 0.4625000059604645,
|
| 62 |
+
"rewards/chosen": -0.2480003386735916,
|
| 63 |
+
"rewards/margins": -0.023857835680246353,
|
| 64 |
+
"rewards/rejected": -0.22414250671863556,
|
| 65 |
+
"step": 40
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.01,
|
| 69 |
+
"learning_rate": 4.72972972972973e-05,
|
| 70 |
+
"logits/chosen": -3.1669116020202637,
|
| 71 |
+
"logits/rejected": -3.1525278091430664,
|
| 72 |
+
"logps/chosen": -123.1195297241211,
|
| 73 |
+
"logps/rejected": -128.38714599609375,
|
| 74 |
+
"loss": 0.6781,
|
| 75 |
+
"rewards/accuracies": 0.6000000238418579,
|
| 76 |
+
"rewards/chosen": -0.19161827862262726,
|
| 77 |
+
"rewards/margins": 0.04476013034582138,
|
| 78 |
+
"rewards/rejected": -0.23637838661670685,
|
| 79 |
+
"step": 50
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.01,
|
| 83 |
+
"learning_rate": 5.694980694980695e-05,
|
| 84 |
+
"logits/chosen": -3.1426501274108887,
|
| 85 |
+
"logits/rejected": -3.132570266723633,
|
| 86 |
+
"logps/chosen": -120.64261627197266,
|
| 87 |
+
"logps/rejected": -113.0268783569336,
|
| 88 |
+
"loss": 0.7107,
|
| 89 |
+
"rewards/accuracies": 0.4749999940395355,
|
| 90 |
+
"rewards/chosen": -0.3047412037849426,
|
| 91 |
+
"rewards/margins": -0.014080168679356575,
|
| 92 |
+
"rewards/rejected": -0.2906610071659088,
|
| 93 |
+
"step": 60
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.01,
|
| 97 |
+
"learning_rate": 6.660231660231661e-05,
|
| 98 |
+
"logits/chosen": -3.170804500579834,
|
| 99 |
+
"logits/rejected": -3.164586305618286,
|
| 100 |
+
"logps/chosen": -116.2149887084961,
|
| 101 |
+
"logps/rejected": -126.68898010253906,
|
| 102 |
+
"loss": 0.6885,
|
| 103 |
+
"rewards/accuracies": 0.5249999761581421,
|
| 104 |
+
"rewards/chosen": -0.33613839745521545,
|
| 105 |
+
"rewards/margins": 0.028235793113708496,
|
| 106 |
+
"rewards/rejected": -0.36437422037124634,
|
| 107 |
+
"step": 70
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.02,
|
| 111 |
+
"learning_rate": 7.625482625482626e-05,
|
| 112 |
+
"logits/chosen": -3.17895245552063,
|
| 113 |
+
"logits/rejected": -3.1590020656585693,
|
| 114 |
+
"logps/chosen": -126.33811950683594,
|
| 115 |
+
"logps/rejected": -103.02183532714844,
|
| 116 |
+
"loss": 0.6215,
|
| 117 |
+
"rewards/accuracies": 0.675000011920929,
|
| 118 |
+
"rewards/chosen": -0.4042009711265564,
|
| 119 |
+
"rewards/margins": 0.19598612189292908,
|
| 120 |
+
"rewards/rejected": -0.6001870632171631,
|
| 121 |
+
"step": 80
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.02,
|
| 125 |
+
"learning_rate": 8.59073359073359e-05,
|
| 126 |
+
"logits/chosen": -3.180785894393921,
|
| 127 |
+
"logits/rejected": -3.1576857566833496,
|
| 128 |
+
"logps/chosen": -129.47866821289062,
|
| 129 |
+
"logps/rejected": -126.99539947509766,
|
| 130 |
+
"loss": 0.6889,
|
| 131 |
+
"rewards/accuracies": 0.5249999761581421,
|
| 132 |
+
"rewards/chosen": -0.8047823905944824,
|
| 133 |
+
"rewards/margins": 0.08432246744632721,
|
| 134 |
+
"rewards/rejected": -0.8891048431396484,
|
| 135 |
+
"step": 90
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.02,
|
| 139 |
+
"learning_rate": 9.555984555984557e-05,
|
| 140 |
+
"logits/chosen": -3.139349937438965,
|
| 141 |
+
"logits/rejected": -3.114441394805908,
|
| 142 |
+
"logps/chosen": -139.87002563476562,
|
| 143 |
+
"logps/rejected": -135.06802368164062,
|
| 144 |
+
"loss": 0.6326,
|
| 145 |
+
"rewards/accuracies": 0.6625000238418579,
|
| 146 |
+
"rewards/chosen": -1.2678108215332031,
|
| 147 |
+
"rewards/margins": 0.2349836528301239,
|
| 148 |
+
"rewards/rejected": -1.5027945041656494,
|
| 149 |
+
"step": 100
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.02,
|
| 153 |
+
"learning_rate": 0.00010424710424710426,
|
| 154 |
+
"logits/chosen": -3.155695676803589,
|
| 155 |
+
"logits/rejected": -3.1192359924316406,
|
| 156 |
+
"logps/chosen": -126.76655578613281,
|
| 157 |
+
"logps/rejected": -119.91800689697266,
|
| 158 |
+
"loss": 0.6925,
|
| 159 |
+
"rewards/accuracies": 0.5874999761581421,
|
| 160 |
+
"rewards/chosen": -1.1118037700653076,
|
| 161 |
+
"rewards/margins": 0.12902173399925232,
|
| 162 |
+
"rewards/rejected": -1.2408255338668823,
|
| 163 |
+
"step": 110
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.02,
|
| 167 |
+
"learning_rate": 0.0001138996138996139,
|
| 168 |
+
"logits/chosen": -3.203996181488037,
|
| 169 |
+
"logits/rejected": -3.181511402130127,
|
| 170 |
+
"logps/chosen": -121.38505554199219,
|
| 171 |
+
"logps/rejected": -123.63691711425781,
|
| 172 |
+
"loss": 0.6905,
|
| 173 |
+
"rewards/accuracies": 0.5625,
|
| 174 |
+
"rewards/chosen": -0.83983314037323,
|
| 175 |
+
"rewards/margins": 0.14224112033843994,
|
| 176 |
+
"rewards/rejected": -0.9820743799209595,
|
| 177 |
+
"step": 120
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.03,
|
| 181 |
+
"learning_rate": 0.00012355212355212355,
|
| 182 |
+
"logits/chosen": -3.199700355529785,
|
| 183 |
+
"logits/rejected": -3.142176628112793,
|
| 184 |
+
"logps/chosen": -144.78311157226562,
|
| 185 |
+
"logps/rejected": -131.46128845214844,
|
| 186 |
+
"loss": 0.7421,
|
| 187 |
+
"rewards/accuracies": 0.550000011920929,
|
| 188 |
+
"rewards/chosen": -0.7906621694564819,
|
| 189 |
+
"rewards/margins": 0.08599194139242172,
|
| 190 |
+
"rewards/rejected": -0.8766541481018066,
|
| 191 |
+
"step": 130
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.03,
|
| 195 |
+
"learning_rate": 0.00013223938223938227,
|
| 196 |
+
"logits/chosen": -3.11432147026062,
|
| 197 |
+
"logits/rejected": -3.0819637775421143,
|
| 198 |
+
"logps/chosen": -131.96109008789062,
|
| 199 |
+
"logps/rejected": -118.2151107788086,
|
| 200 |
+
"loss": 0.7438,
|
| 201 |
+
"rewards/accuracies": 0.5249999761581421,
|
| 202 |
+
"rewards/chosen": -0.9345771074295044,
|
| 203 |
+
"rewards/margins": 0.03206203132867813,
|
| 204 |
+
"rewards/rejected": -0.9666391611099243,
|
| 205 |
+
"step": 140
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.03,
|
| 209 |
+
"learning_rate": 0.00014189189189189188,
|
| 210 |
+
"logits/chosen": -3.103154182434082,
|
| 211 |
+
"logits/rejected": -3.0087850093841553,
|
| 212 |
+
"logps/chosen": -111.38960266113281,
|
| 213 |
+
"logps/rejected": -109.3032455444336,
|
| 214 |
+
"loss": 0.6658,
|
| 215 |
+
"rewards/accuracies": 0.625,
|
| 216 |
+
"rewards/chosen": -0.4318141043186188,
|
| 217 |
+
"rewards/margins": 0.15236088633537292,
|
| 218 |
+
"rewards/rejected": -0.5841749906539917,
|
| 219 |
+
"step": 150
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.03,
|
| 223 |
+
"learning_rate": 0.00015154440154440155,
|
| 224 |
+
"logits/chosen": -3.065882444381714,
|
| 225 |
+
"logits/rejected": -3.014258623123169,
|
| 226 |
+
"logps/chosen": -115.94306945800781,
|
| 227 |
+
"logps/rejected": -129.7348175048828,
|
| 228 |
+
"loss": 0.7302,
|
| 229 |
+
"rewards/accuracies": 0.5625,
|
| 230 |
+
"rewards/chosen": -0.6290556192398071,
|
| 231 |
+
"rewards/margins": 0.06416401267051697,
|
| 232 |
+
"rewards/rejected": -0.6932196021080017,
|
| 233 |
+
"step": 160
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.03,
|
| 237 |
+
"learning_rate": 0.0001611969111969112,
|
| 238 |
+
"logits/chosen": -2.8696858882904053,
|
| 239 |
+
"logits/rejected": -2.820652723312378,
|
| 240 |
+
"logps/chosen": -121.01307678222656,
|
| 241 |
+
"logps/rejected": -122.356201171875,
|
| 242 |
+
"loss": 0.7221,
|
| 243 |
+
"rewards/accuracies": 0.5249999761581421,
|
| 244 |
+
"rewards/chosen": -0.7266199588775635,
|
| 245 |
+
"rewards/margins": 0.14146149158477783,
|
| 246 |
+
"rewards/rejected": -0.8680814504623413,
|
| 247 |
+
"step": 170
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.03,
|
| 251 |
+
"learning_rate": 0.00017084942084942084,
|
| 252 |
+
"logits/chosen": -2.89375638961792,
|
| 253 |
+
"logits/rejected": -2.8223800659179688,
|
| 254 |
+
"logps/chosen": -131.37777709960938,
|
| 255 |
+
"logps/rejected": -125.69004821777344,
|
| 256 |
+
"loss": 0.583,
|
| 257 |
+
"rewards/accuracies": 0.699999988079071,
|
| 258 |
+
"rewards/chosen": -0.9062817692756653,
|
| 259 |
+
"rewards/margins": 0.5299333333969116,
|
| 260 |
+
"rewards/rejected": -1.4362150430679321,
|
| 261 |
+
"step": 180
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.04,
|
| 265 |
+
"learning_rate": 0.0001805019305019305,
|
| 266 |
+
"logits/chosen": -2.8086953163146973,
|
| 267 |
+
"logits/rejected": -2.8810436725616455,
|
| 268 |
+
"logps/chosen": -115.71038818359375,
|
| 269 |
+
"logps/rejected": -133.2216339111328,
|
| 270 |
+
"loss": 0.7593,
|
| 271 |
+
"rewards/accuracies": 0.5625,
|
| 272 |
+
"rewards/chosen": -1.390483021736145,
|
| 273 |
+
"rewards/margins": 0.12736426293849945,
|
| 274 |
+
"rewards/rejected": -1.5178472995758057,
|
| 275 |
+
"step": 190
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.04,
|
| 279 |
+
"learning_rate": 0.00019015444015444015,
|
| 280 |
+
"logits/chosen": -3.0167882442474365,
|
| 281 |
+
"logits/rejected": -2.996938943862915,
|
| 282 |
+
"logps/chosen": -118.67408752441406,
|
| 283 |
+
"logps/rejected": -106.15169525146484,
|
| 284 |
+
"loss": 0.672,
|
| 285 |
+
"rewards/accuracies": 0.5874999761581421,
|
| 286 |
+
"rewards/chosen": -0.7122364640235901,
|
| 287 |
+
"rewards/margins": 0.25746825337409973,
|
| 288 |
+
"rewards/rejected": -0.9697047472000122,
|
| 289 |
+
"step": 200
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.04,
|
| 293 |
+
"eval_logits/chosen": -3.1546952724456787,
|
| 294 |
+
"eval_logits/rejected": -3.1338424682617188,
|
| 295 |
+
"eval_logps/chosen": -127.55575561523438,
|
| 296 |
+
"eval_logps/rejected": -128.7761993408203,
|
| 297 |
+
"eval_loss": 0.743442177772522,
|
| 298 |
+
"eval_rewards/accuracies": 0.5858798623085022,
|
| 299 |
+
"eval_rewards/chosen": -1.0754988193511963,
|
| 300 |
+
"eval_rewards/margins": 0.3047899305820465,
|
| 301 |
+
"eval_rewards/rejected": -1.3802887201309204,
|
| 302 |
+
"eval_runtime": 1335.2681,
|
| 303 |
+
"eval_samples_per_second": 0.711,
|
| 304 |
+
"eval_steps_per_second": 0.711,
|
| 305 |
+
"step": 200
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 0.04,
|
| 309 |
+
"learning_rate": 0.0001998069498069498,
|
| 310 |
+
"logits/chosen": -3.2106406688690186,
|
| 311 |
+
"logits/rejected": -3.1746304035186768,
|
| 312 |
+
"logps/chosen": -131.38128662109375,
|
| 313 |
+
"logps/rejected": -121.36234283447266,
|
| 314 |
+
"loss": 0.664,
|
| 315 |
+
"rewards/accuracies": 0.637499988079071,
|
| 316 |
+
"rewards/chosen": -1.005927324295044,
|
| 317 |
+
"rewards/margins": 0.5132917165756226,
|
| 318 |
+
"rewards/rejected": -1.5192190408706665,
|
| 319 |
+
"step": 210
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 0.04,
|
| 323 |
+
"learning_rate": 0.00020945945945945947,
|
| 324 |
+
"logits/chosen": -3.2851333618164062,
|
| 325 |
+
"logits/rejected": -3.267256259918213,
|
| 326 |
+
"logps/chosen": -120.11392974853516,
|
| 327 |
+
"logps/rejected": -117.89964294433594,
|
| 328 |
+
"loss": 0.6965,
|
| 329 |
+
"rewards/accuracies": 0.612500011920929,
|
| 330 |
+
"rewards/chosen": -0.7865055203437805,
|
| 331 |
+
"rewards/margins": 0.19706687331199646,
|
| 332 |
+
"rewards/rejected": -0.9835723638534546,
|
| 333 |
+
"step": 220
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 0.04,
|
| 337 |
+
"learning_rate": 0.0002191119691119691,
|
| 338 |
+
"logits/chosen": -3.408318281173706,
|
| 339 |
+
"logits/rejected": -3.351940870285034,
|
| 340 |
+
"logps/chosen": -107.76014709472656,
|
| 341 |
+
"logps/rejected": -107.02482604980469,
|
| 342 |
+
"loss": 0.7385,
|
| 343 |
+
"rewards/accuracies": 0.512499988079071,
|
| 344 |
+
"rewards/chosen": -0.41882553696632385,
|
| 345 |
+
"rewards/margins": 0.01567765511572361,
|
| 346 |
+
"rewards/rejected": -0.4345032274723053,
|
| 347 |
+
"step": 230
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 0.05,
|
| 351 |
+
"learning_rate": 0.00022876447876447875,
|
| 352 |
+
"logits/chosen": -3.155478000640869,
|
| 353 |
+
"logits/rejected": -3.135077953338623,
|
| 354 |
+
"logps/chosen": -131.94711303710938,
|
| 355 |
+
"logps/rejected": -112.65836334228516,
|
| 356 |
+
"loss": 0.7328,
|
| 357 |
+
"rewards/accuracies": 0.5375000238418579,
|
| 358 |
+
"rewards/chosen": -0.6660552024841309,
|
| 359 |
+
"rewards/margins": 0.10303208976984024,
|
| 360 |
+
"rewards/rejected": -0.7690872550010681,
|
| 361 |
+
"step": 240
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 0.05,
|
| 365 |
+
"learning_rate": 0.00023841698841698842,
|
| 366 |
+
"logits/chosen": -3.290839433670044,
|
| 367 |
+
"logits/rejected": -3.2837767601013184,
|
| 368 |
+
"logps/chosen": -132.83676147460938,
|
| 369 |
+
"logps/rejected": -119.4383773803711,
|
| 370 |
+
"loss": 0.678,
|
| 371 |
+
"rewards/accuracies": 0.5874999761581421,
|
| 372 |
+
"rewards/chosen": -0.9368747472763062,
|
| 373 |
+
"rewards/margins": 0.25725504755973816,
|
| 374 |
+
"rewards/rejected": -1.1941298246383667,
|
| 375 |
+
"step": 250
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 0.05,
|
| 379 |
+
"learning_rate": 0.0002480694980694981,
|
| 380 |
+
"logits/chosen": -3.3188934326171875,
|
| 381 |
+
"logits/rejected": -3.361811876296997,
|
| 382 |
+
"logps/chosen": -130.7174835205078,
|
| 383 |
+
"logps/rejected": -147.4563751220703,
|
| 384 |
+
"loss": 0.7069,
|
| 385 |
+
"rewards/accuracies": 0.574999988079071,
|
| 386 |
+
"rewards/chosen": -0.9200389981269836,
|
| 387 |
+
"rewards/margins": 0.4393937587738037,
|
| 388 |
+
"rewards/rejected": -1.3594326972961426,
|
| 389 |
+
"step": 260
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 0.05,
|
| 393 |
+
"learning_rate": 0.0002567567567567567,
|
| 394 |
+
"logits/chosen": -3.249516010284424,
|
| 395 |
+
"logits/rejected": -3.2256407737731934,
|
| 396 |
+
"logps/chosen": -108.8515853881836,
|
| 397 |
+
"logps/rejected": -132.24813842773438,
|
| 398 |
+
"loss": 0.6983,
|
| 399 |
+
"rewards/accuracies": 0.612500011920929,
|
| 400 |
+
"rewards/chosen": -0.42829591035842896,
|
| 401 |
+
"rewards/margins": 0.3723019063472748,
|
| 402 |
+
"rewards/rejected": -0.8005977869033813,
|
| 403 |
+
"step": 270
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 0.05,
|
| 407 |
+
"learning_rate": 0.0002635135135135135,
|
| 408 |
+
"logits/chosen": -3.0320539474487305,
|
| 409 |
+
"logits/rejected": -2.982640027999878,
|
| 410 |
+
"logps/chosen": -175.3157501220703,
|
| 411 |
+
"logps/rejected": -154.48963928222656,
|
| 412 |
+
"loss": 1.4131,
|
| 413 |
+
"rewards/accuracies": 0.512499988079071,
|
| 414 |
+
"rewards/chosen": -2.973146915435791,
|
| 415 |
+
"rewards/margins": -0.0525052547454834,
|
| 416 |
+
"rewards/rejected": -2.9206414222717285,
|
| 417 |
+
"step": 280
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 0.06,
|
| 421 |
+
"learning_rate": 0.00027316602316602317,
|
| 422 |
+
"logits/chosen": -2.9082584381103516,
|
| 423 |
+
"logits/rejected": -2.8451313972473145,
|
| 424 |
+
"logps/chosen": -129.400146484375,
|
| 425 |
+
"logps/rejected": -118.70863342285156,
|
| 426 |
+
"loss": 0.9237,
|
| 427 |
+
"rewards/accuracies": 0.48750001192092896,
|
| 428 |
+
"rewards/chosen": -1.8987582921981812,
|
| 429 |
+
"rewards/margins": 0.024892251938581467,
|
| 430 |
+
"rewards/rejected": -1.9236505031585693,
|
| 431 |
+
"step": 290
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 0.06,
|
| 435 |
+
"learning_rate": 0.00028281853281853284,
|
| 436 |
+
"logits/chosen": -3.066049337387085,
|
| 437 |
+
"logits/rejected": -2.9455220699310303,
|
| 438 |
+
"logps/chosen": -131.84317016601562,
|
| 439 |
+
"logps/rejected": -121.24568176269531,
|
| 440 |
+
"loss": 0.6998,
|
| 441 |
+
"rewards/accuracies": 0.574999988079071,
|
| 442 |
+
"rewards/chosen": -1.1568044424057007,
|
| 443 |
+
"rewards/margins": 0.3450776934623718,
|
| 444 |
+
"rewards/rejected": -1.5018823146820068,
|
| 445 |
+
"step": 300
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 0.06,
|
| 449 |
+
"learning_rate": 0.0002924710424710425,
|
| 450 |
+
"logits/chosen": -2.8079309463500977,
|
| 451 |
+
"logits/rejected": -2.773160457611084,
|
| 452 |
+
"logps/chosen": -169.48828125,
|
| 453 |
+
"logps/rejected": -148.60374450683594,
|
| 454 |
+
"loss": 2.3505,
|
| 455 |
+
"rewards/accuracies": 0.5375000238418579,
|
| 456 |
+
"rewards/chosen": -5.812082290649414,
|
| 457 |
+
"rewards/margins": -0.6856533288955688,
|
| 458 |
+
"rewards/rejected": -5.126428127288818,
|
| 459 |
+
"step": 310
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 0.06,
|
| 463 |
+
"learning_rate": 0.0003021235521235521,
|
| 464 |
+
"logits/chosen": -2.6579480171203613,
|
| 465 |
+
"logits/rejected": -2.6659698486328125,
|
| 466 |
+
"logps/chosen": -141.78701782226562,
|
| 467 |
+
"logps/rejected": -160.35110473632812,
|
| 468 |
+
"loss": 0.8982,
|
| 469 |
+
"rewards/accuracies": 0.4749999940395355,
|
| 470 |
+
"rewards/chosen": -3.301175594329834,
|
| 471 |
+
"rewards/margins": 0.01409349124878645,
|
| 472 |
+
"rewards/rejected": -3.3152689933776855,
|
| 473 |
+
"step": 320
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 0.06,
|
| 477 |
+
"learning_rate": 0.0003117760617760618,
|
| 478 |
+
"logits/chosen": -2.8775956630706787,
|
| 479 |
+
"logits/rejected": -2.8255507946014404,
|
| 480 |
+
"logps/chosen": -150.46177673339844,
|
| 481 |
+
"logps/rejected": -144.97073364257812,
|
| 482 |
+
"loss": 0.9436,
|
| 483 |
+
"rewards/accuracies": 0.48750001192092896,
|
| 484 |
+
"rewards/chosen": -3.29494047164917,
|
| 485 |
+
"rewards/margins": 0.0633418932557106,
|
| 486 |
+
"rewards/rejected": -3.3582825660705566,
|
| 487 |
+
"step": 330
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 0.07,
|
| 491 |
+
"learning_rate": 0.0003204633204633205,
|
| 492 |
+
"logits/chosen": -2.8166918754577637,
|
| 493 |
+
"logits/rejected": -2.818556308746338,
|
| 494 |
+
"logps/chosen": -171.34437561035156,
|
| 495 |
+
"logps/rejected": -172.5870361328125,
|
| 496 |
+
"loss": 1.0895,
|
| 497 |
+
"rewards/accuracies": 0.574999988079071,
|
| 498 |
+
"rewards/chosen": -4.722414016723633,
|
| 499 |
+
"rewards/margins": 0.23760518431663513,
|
| 500 |
+
"rewards/rejected": -4.960019111633301,
|
| 501 |
+
"step": 340
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 0.07,
|
| 505 |
+
"learning_rate": 0.00033011583011583015,
|
| 506 |
+
"logits/chosen": -2.9005284309387207,
|
| 507 |
+
"logits/rejected": -2.905046224594116,
|
| 508 |
+
"logps/chosen": -179.5493621826172,
|
| 509 |
+
"logps/rejected": -189.4880828857422,
|
| 510 |
+
"loss": 1.4019,
|
| 511 |
+
"rewards/accuracies": 0.512499988079071,
|
| 512 |
+
"rewards/chosen": -4.879184722900391,
|
| 513 |
+
"rewards/margins": 0.2637065649032593,
|
| 514 |
+
"rewards/rejected": -5.142890930175781,
|
| 515 |
+
"step": 350
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 0.07,
|
| 519 |
+
"learning_rate": 0.00033976833976833977,
|
| 520 |
+
"logits/chosen": -2.4955551624298096,
|
| 521 |
+
"logits/rejected": -2.5379605293273926,
|
| 522 |
+
"logps/chosen": -142.00270080566406,
|
| 523 |
+
"logps/rejected": -147.41220092773438,
|
| 524 |
+
"loss": 0.7586,
|
| 525 |
+
"rewards/accuracies": 0.625,
|
| 526 |
+
"rewards/chosen": -3.0069820880889893,
|
| 527 |
+
"rewards/margins": 0.49537187814712524,
|
| 528 |
+
"rewards/rejected": -3.502354383468628,
|
| 529 |
+
"step": 360
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 0.07,
|
| 533 |
+
"learning_rate": 0.00034942084942084944,
|
| 534 |
+
"logits/chosen": -2.451601266860962,
|
| 535 |
+
"logits/rejected": -2.4406824111938477,
|
| 536 |
+
"logps/chosen": -177.74746704101562,
|
| 537 |
+
"logps/rejected": -181.7180633544922,
|
| 538 |
+
"loss": 1.6042,
|
| 539 |
+
"rewards/accuracies": 0.512499988079071,
|
| 540 |
+
"rewards/chosen": -6.1398024559021,
|
| 541 |
+
"rewards/margins": -0.07387089729309082,
|
| 542 |
+
"rewards/rejected": -6.065931797027588,
|
| 543 |
+
"step": 370
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 0.07,
|
| 547 |
+
"learning_rate": 0.0003590733590733591,
|
| 548 |
+
"logits/chosen": -2.4471051692962646,
|
| 549 |
+
"logits/rejected": -2.409393548965454,
|
| 550 |
+
"logps/chosen": -182.06051635742188,
|
| 551 |
+
"logps/rejected": -163.97035217285156,
|
| 552 |
+
"loss": 2.4314,
|
| 553 |
+
"rewards/accuracies": 0.4375,
|
| 554 |
+
"rewards/chosen": -6.733994483947754,
|
| 555 |
+
"rewards/margins": -1.0469824075698853,
|
| 556 |
+
"rewards/rejected": -5.687012672424316,
|
| 557 |
+
"step": 380
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 0.08,
|
| 561 |
+
"learning_rate": 0.0003687258687258687,
|
| 562 |
+
"logits/chosen": -2.8171439170837402,
|
| 563 |
+
"logits/rejected": -2.7251698970794678,
|
| 564 |
+
"logps/chosen": -160.04652404785156,
|
| 565 |
+
"logps/rejected": -142.2588348388672,
|
| 566 |
+
"loss": 1.0093,
|
| 567 |
+
"rewards/accuracies": 0.574999988079071,
|
| 568 |
+
"rewards/chosen": -3.551713466644287,
|
| 569 |
+
"rewards/margins": 0.07490300387144089,
|
| 570 |
+
"rewards/rejected": -3.6266167163848877,
|
| 571 |
+
"step": 390
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 0.08,
|
| 575 |
+
"learning_rate": 0.0003783783783783784,
|
| 576 |
+
"logits/chosen": -2.8222527503967285,
|
| 577 |
+
"logits/rejected": -2.8754923343658447,
|
| 578 |
+
"logps/chosen": -138.70736694335938,
|
| 579 |
+
"logps/rejected": -142.30128479003906,
|
| 580 |
+
"loss": 0.945,
|
| 581 |
+
"rewards/accuracies": 0.512499988079071,
|
| 582 |
+
"rewards/chosen": -3.170539140701294,
|
| 583 |
+
"rewards/margins": 0.011356920003890991,
|
| 584 |
+
"rewards/rejected": -3.1818957328796387,
|
| 585 |
+
"step": 400
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 0.08,
|
| 589 |
+
"eval_logits/chosen": -2.9538896083831787,
|
| 590 |
+
"eval_logits/rejected": -2.8971762657165527,
|
| 591 |
+
"eval_logps/chosen": -152.2897491455078,
|
| 592 |
+
"eval_logps/rejected": -150.13941955566406,
|
| 593 |
+
"eval_loss": 1.0823436975479126,
|
| 594 |
+
"eval_rewards/accuracies": 0.5100105404853821,
|
| 595 |
+
"eval_rewards/chosen": -3.548898458480835,
|
| 596 |
+
"eval_rewards/margins": -0.032288454473018646,
|
| 597 |
+
"eval_rewards/rejected": -3.5166099071502686,
|
| 598 |
+
"eval_runtime": 1347.2142,
|
| 599 |
+
"eval_samples_per_second": 0.704,
|
| 600 |
+
"eval_steps_per_second": 0.704,
|
| 601 |
+
"step": 400
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 0.08,
|
| 605 |
+
"learning_rate": 0.00038803088803088807,
|
| 606 |
+
"logits/chosen": -2.8151559829711914,
|
| 607 |
+
"logits/rejected": -2.8057456016540527,
|
| 608 |
+
"logps/chosen": -145.69772338867188,
|
| 609 |
+
"logps/rejected": -156.96774291992188,
|
| 610 |
+
"loss": 0.8857,
|
| 611 |
+
"rewards/accuracies": 0.574999988079071,
|
| 612 |
+
"rewards/chosen": -3.8467044830322266,
|
| 613 |
+
"rewards/margins": 0.3242764472961426,
|
| 614 |
+
"rewards/rejected": -4.170981407165527,
|
| 615 |
+
"step": 410
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 0.08,
|
| 619 |
+
"learning_rate": 0.0003976833976833977,
|
| 620 |
+
"logits/chosen": -2.7274651527404785,
|
| 621 |
+
"logits/rejected": -2.713927745819092,
|
| 622 |
+
"logps/chosen": -164.06692504882812,
|
| 623 |
+
"logps/rejected": -149.8723907470703,
|
| 624 |
+
"loss": 0.9745,
|
| 625 |
+
"rewards/accuracies": 0.512499988079071,
|
| 626 |
+
"rewards/chosen": -3.5636115074157715,
|
| 627 |
+
"rewards/margins": 0.0345739908516407,
|
| 628 |
+
"rewards/rejected": -3.5981857776641846,
|
| 629 |
+
"step": 420
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 0.08,
|
| 633 |
+
"learning_rate": 0.00040733590733590735,
|
| 634 |
+
"logits/chosen": -2.8782758712768555,
|
| 635 |
+
"logits/rejected": -2.8031697273254395,
|
| 636 |
+
"logps/chosen": -165.91473388671875,
|
| 637 |
+
"logps/rejected": -147.03627014160156,
|
| 638 |
+
"loss": 0.8281,
|
| 639 |
+
"rewards/accuracies": 0.625,
|
| 640 |
+
"rewards/chosen": -4.2417216300964355,
|
| 641 |
+
"rewards/margins": 0.32142549753189087,
|
| 642 |
+
"rewards/rejected": -4.563147068023682,
|
| 643 |
+
"step": 430
|
| 644 |
+
},
|
| 645 |
+
{
|
| 646 |
+
"epoch": 0.09,
|
| 647 |
+
"learning_rate": 0.000416988416988417,
|
| 648 |
+
"logits/chosen": -2.623711585998535,
|
| 649 |
+
"logits/rejected": -2.622528314590454,
|
| 650 |
+
"logps/chosen": -149.8426513671875,
|
| 651 |
+
"logps/rejected": -159.93692016601562,
|
| 652 |
+
"loss": 0.9961,
|
| 653 |
+
"rewards/accuracies": 0.5,
|
| 654 |
+
"rewards/chosen": -4.433084487915039,
|
| 655 |
+
"rewards/margins": 0.28495556116104126,
|
| 656 |
+
"rewards/rejected": -4.7180399894714355,
|
| 657 |
+
"step": 440
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 0.09,
|
| 661 |
+
"learning_rate": 0.00042664092664092664,
|
| 662 |
+
"logits/chosen": -2.580371618270874,
|
| 663 |
+
"logits/rejected": -2.5741231441497803,
|
| 664 |
+
"logps/chosen": -160.84347534179688,
|
| 665 |
+
"logps/rejected": -141.23475646972656,
|
| 666 |
+
"loss": 1.2914,
|
| 667 |
+
"rewards/accuracies": 0.4625000059604645,
|
| 668 |
+
"rewards/chosen": -4.076364994049072,
|
| 669 |
+
"rewards/margins": -0.23183032870292664,
|
| 670 |
+
"rewards/rejected": -3.8445351123809814,
|
| 671 |
+
"step": 450
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 0.09,
|
| 675 |
+
"learning_rate": 0.0004362934362934363,
|
| 676 |
+
"logits/chosen": -2.889563798904419,
|
| 677 |
+
"logits/rejected": -2.8342082500457764,
|
| 678 |
+
"logps/chosen": -186.77017211914062,
|
| 679 |
+
"logps/rejected": -168.42330932617188,
|
| 680 |
+
"loss": 1.1317,
|
| 681 |
+
"rewards/accuracies": 0.44999998807907104,
|
| 682 |
+
"rewards/chosen": -4.674792289733887,
|
| 683 |
+
"rewards/margins": -0.265504390001297,
|
| 684 |
+
"rewards/rejected": -4.409287929534912,
|
| 685 |
+
"step": 460
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 0.09,
|
| 689 |
+
"learning_rate": 0.000445945945945946,
|
| 690 |
+
"logits/chosen": -2.3731606006622314,
|
| 691 |
+
"logits/rejected": -2.344404697418213,
|
| 692 |
+
"logps/chosen": -172.8909454345703,
|
| 693 |
+
"logps/rejected": -175.9696502685547,
|
| 694 |
+
"loss": 0.9674,
|
| 695 |
+
"rewards/accuracies": 0.574999988079071,
|
| 696 |
+
"rewards/chosen": -5.584943771362305,
|
| 697 |
+
"rewards/margins": 0.14772634208202362,
|
| 698 |
+
"rewards/rejected": -5.732670783996582,
|
| 699 |
+
"step": 470
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 0.09,
|
| 703 |
+
"learning_rate": 0.0004555984555984556,
|
| 704 |
+
"logits/chosen": -2.3436319828033447,
|
| 705 |
+
"logits/rejected": -2.301845073699951,
|
| 706 |
+
"logps/chosen": -173.07313537597656,
|
| 707 |
+
"logps/rejected": -169.7339630126953,
|
| 708 |
+
"loss": 1.2097,
|
| 709 |
+
"rewards/accuracies": 0.4749999940395355,
|
| 710 |
+
"rewards/chosen": -5.504385471343994,
|
| 711 |
+
"rewards/margins": 0.23244301974773407,
|
| 712 |
+
"rewards/rejected": -5.736828804016113,
|
| 713 |
+
"step": 480
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 0.09,
|
| 717 |
+
"learning_rate": 0.00046525096525096526,
|
| 718 |
+
"logits/chosen": -2.6778111457824707,
|
| 719 |
+
"logits/rejected": -2.5807526111602783,
|
| 720 |
+
"logps/chosen": -166.180419921875,
|
| 721 |
+
"logps/rejected": -168.2677001953125,
|
| 722 |
+
"loss": 1.2799,
|
| 723 |
+
"rewards/accuracies": 0.512499988079071,
|
| 724 |
+
"rewards/chosen": -4.63643217086792,
|
| 725 |
+
"rewards/margins": 0.5320998430252075,
|
| 726 |
+
"rewards/rejected": -5.168532371520996,
|
| 727 |
+
"step": 490
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 0.1,
|
| 731 |
+
"learning_rate": 0.00047490347490347493,
|
| 732 |
+
"logits/chosen": -3.083743095397949,
|
| 733 |
+
"logits/rejected": -3.080765962600708,
|
| 734 |
+
"logps/chosen": -193.96841430664062,
|
| 735 |
+
"logps/rejected": -182.3042755126953,
|
| 736 |
+
"loss": 1.6246,
|
| 737 |
+
"rewards/accuracies": 0.5,
|
| 738 |
+
"rewards/chosen": -6.152979850769043,
|
| 739 |
+
"rewards/margins": -0.20669928193092346,
|
| 740 |
+
"rewards/rejected": -5.946280479431152,
|
| 741 |
+
"step": 500
|
| 742 |
+
}
|
| 743 |
+
],
|
| 744 |
+
"logging_steps": 10,
|
| 745 |
+
"max_steps": 5173,
|
| 746 |
+
"num_train_epochs": 1,
|
| 747 |
+
"save_steps": 100,
|
| 748 |
+
"total_flos": 0.0,
|
| 749 |
+
"trial_name": null,
|
| 750 |
+
"trial_params": null
|
| 751 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b2d24f52f3de130e83c614d71d9b7667bb85da1de04a98bd39d203115dbe7af7
|
| 3 |
+
size 4091
|