Upload folder using huggingface_hub
#1
by
sameepv21
- opened
- README.md +9 -0
- adapter_config.json +26 -0
- adapter_model.bin +3 -0
- config.json +46 -0
- non_lora_trainables.bin +3 -0
- trainer_state.json +846 -0
README.md
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: peft
|
| 3 |
+
---
|
| 4 |
+
## Training procedure
|
| 5 |
+
|
| 6 |
+
### Framework versions
|
| 7 |
+
|
| 8 |
+
|
| 9 |
+
- PEFT 0.5.0
|
adapter_config.json
ADDED
|
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"auto_mapping": null,
|
| 3 |
+
"base_model_name_or_path": "ShareGPTVideo/LLaVA-Hound-SFT",
|
| 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": 256,
|
| 11 |
+
"lora_dropout": 0.05,
|
| 12 |
+
"modules_to_save": null,
|
| 13 |
+
"peft_type": "LORA",
|
| 14 |
+
"r": 128,
|
| 15 |
+
"revision": null,
|
| 16 |
+
"target_modules": [
|
| 17 |
+
"v_proj",
|
| 18 |
+
"k_proj",
|
| 19 |
+
"gate_proj",
|
| 20 |
+
"q_proj",
|
| 21 |
+
"up_proj",
|
| 22 |
+
"o_proj",
|
| 23 |
+
"down_proj"
|
| 24 |
+
],
|
| 25 |
+
"task_type": "CAUSAL_LM"
|
| 26 |
+
}
|
adapter_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:140feae11930e5f4f9fcad29eb7eecb2f7bd6b38a4b2d8c25c9a89e4d098d1be
|
| 3 |
+
size 639787082
|
config.json
ADDED
|
@@ -0,0 +1,46 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"X": [
|
| 3 |
+
"Image",
|
| 4 |
+
"Video"
|
| 5 |
+
],
|
| 6 |
+
"_name_or_path": "ShareGPTVideo/LLaVA-Hound-SFT",
|
| 7 |
+
"architectures": [
|
| 8 |
+
"LlavaLlamaForCausalLM"
|
| 9 |
+
],
|
| 10 |
+
"attention_bias": false,
|
| 11 |
+
"attention_dropout": 0.0,
|
| 12 |
+
"bos_token_id": 1,
|
| 13 |
+
"eos_token_id": 2,
|
| 14 |
+
"freeze_mm_mlp_adapter": true,
|
| 15 |
+
"hidden_act": "silu",
|
| 16 |
+
"hidden_size": 4096,
|
| 17 |
+
"image_aspect_ratio": "pad",
|
| 18 |
+
"image_grid_pinpoints": null,
|
| 19 |
+
"initializer_range": 0.02,
|
| 20 |
+
"intermediate_size": 11008,
|
| 21 |
+
"max_position_embeddings": 4096,
|
| 22 |
+
"mm_hidden_size": 1024,
|
| 23 |
+
"mm_image_tower": "LanguageBind/LanguageBind_Image",
|
| 24 |
+
"mm_projector_type": "mlp2x_gelu",
|
| 25 |
+
"mm_use_x_patch_token": false,
|
| 26 |
+
"mm_use_x_start_end": false,
|
| 27 |
+
"mm_video_tower": "LanguageBind/LanguageBind_Video_merge",
|
| 28 |
+
"mm_vision_select_feature": "patch",
|
| 29 |
+
"mm_vision_select_layer": -2,
|
| 30 |
+
"model_type": "llava_llama",
|
| 31 |
+
"num_attention_heads": 32,
|
| 32 |
+
"num_hidden_layers": 32,
|
| 33 |
+
"num_key_value_heads": 32,
|
| 34 |
+
"pad_token_id": 0,
|
| 35 |
+
"pretraining_tp": 1,
|
| 36 |
+
"rms_norm_eps": 1e-05,
|
| 37 |
+
"rope_scaling": null,
|
| 38 |
+
"rope_theta": 10000.0,
|
| 39 |
+
"tie_word_embeddings": false,
|
| 40 |
+
"torch_dtype": "bfloat16",
|
| 41 |
+
"transformers_version": "4.37.0",
|
| 42 |
+
"tune_mm_mlp_adapter": false,
|
| 43 |
+
"use_cache": true,
|
| 44 |
+
"use_mm_proj": true,
|
| 45 |
+
"vocab_size": 32000
|
| 46 |
+
}
|
non_lora_trainables.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:60fb82c3660319e6d0b239950b20c28181e97f1ade117dc0660b40e2ad94a89b
|
| 3 |
+
size 912
|
trainer_state.json
ADDED
|
@@ -0,0 +1,846 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 1200,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.06,
|
| 13 |
+
"learning_rate": 1.0416666666666667e-07,
|
| 14 |
+
"logps/chosen": -46.291481018066406,
|
| 15 |
+
"logps/rejected": -63.38591384887695,
|
| 16 |
+
"loss": 0.6933,
|
| 17 |
+
"losses/dpo": 0.6924217939376831,
|
| 18 |
+
"losses/sft": 1.2695705890655518,
|
| 19 |
+
"losses/total": 0.6924217939376831,
|
| 20 |
+
"ref_logps/chosen": -46.2892951965332,
|
| 21 |
+
"ref_logps/rejected": -63.38469314575195,
|
| 22 |
+
"rewards/accuracies": 0.46150001883506775,
|
| 23 |
+
"rewards/chosen": -0.00021846062736585736,
|
| 24 |
+
"rewards/margins": -9.587412932887673e-05,
|
| 25 |
+
"rewards/rejected": -0.00012258654169272631,
|
| 26 |
+
"step": 25
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.12,
|
| 30 |
+
"learning_rate": 2.0833333333333333e-07,
|
| 31 |
+
"logps/chosen": -44.6225471496582,
|
| 32 |
+
"logps/rejected": -59.81039047241211,
|
| 33 |
+
"loss": 0.6932,
|
| 34 |
+
"losses/dpo": 0.6935679912567139,
|
| 35 |
+
"losses/sft": 1.2453413009643555,
|
| 36 |
+
"losses/total": 0.6935679912567139,
|
| 37 |
+
"ref_logps/chosen": -44.617828369140625,
|
| 38 |
+
"ref_logps/rejected": -59.8062744140625,
|
| 39 |
+
"rewards/accuracies": 0.503000020980835,
|
| 40 |
+
"rewards/chosen": -0.00047206657472997904,
|
| 41 |
+
"rewards/margins": -6.079748345655389e-05,
|
| 42 |
+
"rewards/rejected": -0.00041126925498247147,
|
| 43 |
+
"step": 50
|
| 44 |
+
},
|
| 45 |
+
{
|
| 46 |
+
"epoch": 0.19,
|
| 47 |
+
"learning_rate": 3.1249999999999997e-07,
|
| 48 |
+
"logps/chosen": -43.78951644897461,
|
| 49 |
+
"logps/rejected": -59.19944381713867,
|
| 50 |
+
"loss": 0.6921,
|
| 51 |
+
"losses/dpo": 0.6920485496520996,
|
| 52 |
+
"losses/sft": 1.2684202194213867,
|
| 53 |
+
"losses/total": 0.6920485496520996,
|
| 54 |
+
"ref_logps/chosen": -43.78016662597656,
|
| 55 |
+
"ref_logps/rejected": -59.168540954589844,
|
| 56 |
+
"rewards/accuracies": 0.5205000042915344,
|
| 57 |
+
"rewards/chosen": -0.0009349272586405277,
|
| 58 |
+
"rewards/margins": 0.00215632701292634,
|
| 59 |
+
"rewards/rejected": -0.003091254271566868,
|
| 60 |
+
"step": 75
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 0.25,
|
| 64 |
+
"learning_rate": 4.1666666666666667e-07,
|
| 65 |
+
"logps/chosen": -44.4240837097168,
|
| 66 |
+
"logps/rejected": -58.47013854980469,
|
| 67 |
+
"loss": 0.6901,
|
| 68 |
+
"losses/dpo": 0.6897304654121399,
|
| 69 |
+
"losses/sft": 1.2445145845413208,
|
| 70 |
+
"losses/total": 0.6897304654121399,
|
| 71 |
+
"ref_logps/chosen": -44.39784622192383,
|
| 72 |
+
"ref_logps/rejected": -58.38094711303711,
|
| 73 |
+
"rewards/accuracies": 0.5600000619888306,
|
| 74 |
+
"rewards/chosen": -0.0026234728284180164,
|
| 75 |
+
"rewards/margins": 0.0062958355993032455,
|
| 76 |
+
"rewards/rejected": -0.0089193070307374,
|
| 77 |
+
"step": 100
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 0.31,
|
| 81 |
+
"learning_rate": 4.976851851851852e-07,
|
| 82 |
+
"logps/chosen": -44.144630432128906,
|
| 83 |
+
"logps/rejected": -60.08927917480469,
|
| 84 |
+
"loss": 0.6856,
|
| 85 |
+
"losses/dpo": 0.6853728294372559,
|
| 86 |
+
"losses/sft": 1.209934115409851,
|
| 87 |
+
"losses/total": 0.6853728294372559,
|
| 88 |
+
"ref_logps/chosen": -44.078399658203125,
|
| 89 |
+
"ref_logps/rejected": -59.86705017089844,
|
| 90 |
+
"rewards/accuracies": 0.5929999947547913,
|
| 91 |
+
"rewards/chosen": -0.006622872781008482,
|
| 92 |
+
"rewards/margins": 0.015600004233419895,
|
| 93 |
+
"rewards/rejected": -0.02222287654876709,
|
| 94 |
+
"step": 125
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.38,
|
| 98 |
+
"learning_rate": 4.861111111111111e-07,
|
| 99 |
+
"logps/chosen": -45.84950256347656,
|
| 100 |
+
"logps/rejected": -62.66447830200195,
|
| 101 |
+
"loss": 0.6775,
|
| 102 |
+
"losses/dpo": 0.6779772043228149,
|
| 103 |
+
"losses/sft": 1.281442403793335,
|
| 104 |
+
"losses/total": 0.6779772043228149,
|
| 105 |
+
"ref_logps/chosen": -45.6851921081543,
|
| 106 |
+
"ref_logps/rejected": -62.169559478759766,
|
| 107 |
+
"rewards/accuracies": 0.6209999918937683,
|
| 108 |
+
"rewards/chosen": -0.01643071323633194,
|
| 109 |
+
"rewards/margins": 0.033061932772397995,
|
| 110 |
+
"rewards/rejected": -0.049492646008729935,
|
| 111 |
+
"step": 150
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.44,
|
| 115 |
+
"learning_rate": 4.74537037037037e-07,
|
| 116 |
+
"logps/chosen": -44.53513717651367,
|
| 117 |
+
"logps/rejected": -61.487178802490234,
|
| 118 |
+
"loss": 0.6684,
|
| 119 |
+
"losses/dpo": 0.6683259010314941,
|
| 120 |
+
"losses/sft": 1.2735780477523804,
|
| 121 |
+
"losses/total": 0.6683259010314941,
|
| 122 |
+
"ref_logps/chosen": -44.296630859375,
|
| 123 |
+
"ref_logps/rejected": -60.71174240112305,
|
| 124 |
+
"rewards/accuracies": 0.6439999938011169,
|
| 125 |
+
"rewards/chosen": -0.02384989708662033,
|
| 126 |
+
"rewards/margins": 0.053693462163209915,
|
| 127 |
+
"rewards/rejected": -0.07754335552453995,
|
| 128 |
+
"step": 175
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.5,
|
| 132 |
+
"learning_rate": 4.6296296296296297e-07,
|
| 133 |
+
"logps/chosen": -46.467018127441406,
|
| 134 |
+
"logps/rejected": -63.491580963134766,
|
| 135 |
+
"loss": 0.6563,
|
| 136 |
+
"losses/dpo": 0.6521183252334595,
|
| 137 |
+
"losses/sft": 1.2733054161071777,
|
| 138 |
+
"losses/total": 0.6521183252334595,
|
| 139 |
+
"ref_logps/chosen": -46.05796432495117,
|
| 140 |
+
"ref_logps/rejected": -62.249488830566406,
|
| 141 |
+
"rewards/accuracies": 0.6285000443458557,
|
| 142 |
+
"rewards/chosen": -0.040905579924583435,
|
| 143 |
+
"rewards/margins": 0.0833037793636322,
|
| 144 |
+
"rewards/rejected": -0.12420935928821564,
|
| 145 |
+
"step": 200
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 0.56,
|
| 149 |
+
"learning_rate": 4.513888888888889e-07,
|
| 150 |
+
"logps/chosen": -46.178504943847656,
|
| 151 |
+
"logps/rejected": -64.19964599609375,
|
| 152 |
+
"loss": 0.6429,
|
| 153 |
+
"losses/dpo": 0.6435222029685974,
|
| 154 |
+
"losses/sft": 1.2407046556472778,
|
| 155 |
+
"losses/total": 0.6435222029685974,
|
| 156 |
+
"ref_logps/chosen": -45.60074234008789,
|
| 157 |
+
"ref_logps/rejected": -62.432167053222656,
|
| 158 |
+
"rewards/accuracies": 0.6385000348091125,
|
| 159 |
+
"rewards/chosen": -0.057775672525167465,
|
| 160 |
+
"rewards/margins": 0.11897158622741699,
|
| 161 |
+
"rewards/rejected": -0.17674726247787476,
|
| 162 |
+
"step": 225
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 0.62,
|
| 166 |
+
"learning_rate": 4.398148148148148e-07,
|
| 167 |
+
"logps/chosen": -46.30465316772461,
|
| 168 |
+
"logps/rejected": -62.2493782043457,
|
| 169 |
+
"loss": 0.6363,
|
| 170 |
+
"losses/dpo": 0.634566605091095,
|
| 171 |
+
"losses/sft": 1.2551387548446655,
|
| 172 |
+
"losses/total": 0.634566605091095,
|
| 173 |
+
"ref_logps/chosen": -45.55400085449219,
|
| 174 |
+
"ref_logps/rejected": -60.08576583862305,
|
| 175 |
+
"rewards/accuracies": 0.6335000395774841,
|
| 176 |
+
"rewards/chosen": -0.07506560534238815,
|
| 177 |
+
"rewards/margins": 0.1412954032421112,
|
| 178 |
+
"rewards/rejected": -0.21636100113391876,
|
| 179 |
+
"step": 250
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 0.69,
|
| 183 |
+
"learning_rate": 4.2824074074074075e-07,
|
| 184 |
+
"logps/chosen": -45.26353073120117,
|
| 185 |
+
"logps/rejected": -62.119075775146484,
|
| 186 |
+
"loss": 0.6234,
|
| 187 |
+
"losses/dpo": 0.6106441020965576,
|
| 188 |
+
"losses/sft": 1.3094475269317627,
|
| 189 |
+
"losses/total": 0.6106441020965576,
|
| 190 |
+
"ref_logps/chosen": -44.300174713134766,
|
| 191 |
+
"ref_logps/rejected": -59.338504791259766,
|
| 192 |
+
"rewards/accuracies": 0.6434999704360962,
|
| 193 |
+
"rewards/chosen": -0.09633561968803406,
|
| 194 |
+
"rewards/margins": 0.1817215383052826,
|
| 195 |
+
"rewards/rejected": -0.27805712819099426,
|
| 196 |
+
"step": 275
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 0.75,
|
| 200 |
+
"learning_rate": 4.1666666666666667e-07,
|
| 201 |
+
"logps/chosen": -45.76985168457031,
|
| 202 |
+
"logps/rejected": -63.50913619995117,
|
| 203 |
+
"loss": 0.613,
|
| 204 |
+
"losses/dpo": 0.6014246940612793,
|
| 205 |
+
"losses/sft": 1.2986456155776978,
|
| 206 |
+
"losses/total": 0.6014246940612793,
|
| 207 |
+
"ref_logps/chosen": -44.60796356201172,
|
| 208 |
+
"ref_logps/rejected": -60.128543853759766,
|
| 209 |
+
"rewards/accuracies": 0.6544999480247498,
|
| 210 |
+
"rewards/chosen": -0.11618894338607788,
|
| 211 |
+
"rewards/margins": 0.2218700349330902,
|
| 212 |
+
"rewards/rejected": -0.3380589783191681,
|
| 213 |
+
"step": 300
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.81,
|
| 217 |
+
"learning_rate": 4.050925925925926e-07,
|
| 218 |
+
"logps/chosen": -46.350311279296875,
|
| 219 |
+
"logps/rejected": -64.47545623779297,
|
| 220 |
+
"loss": 0.6045,
|
| 221 |
+
"losses/dpo": 0.6078605055809021,
|
| 222 |
+
"losses/sft": 1.2730615139007568,
|
| 223 |
+
"losses/total": 0.6078605055809021,
|
| 224 |
+
"ref_logps/chosen": -44.98001480102539,
|
| 225 |
+
"ref_logps/rejected": -60.557586669921875,
|
| 226 |
+
"rewards/accuracies": 0.6600000262260437,
|
| 227 |
+
"rewards/chosen": -0.13702912628650665,
|
| 228 |
+
"rewards/margins": 0.25475841760635376,
|
| 229 |
+
"rewards/rejected": -0.3917875587940216,
|
| 230 |
+
"step": 325
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.88,
|
| 234 |
+
"learning_rate": 3.9351851851851854e-07,
|
| 235 |
+
"logps/chosen": -46.390987396240234,
|
| 236 |
+
"logps/rejected": -65.16218566894531,
|
| 237 |
+
"loss": 0.5944,
|
| 238 |
+
"losses/dpo": 0.5973597168922424,
|
| 239 |
+
"losses/sft": 1.3051469326019287,
|
| 240 |
+
"losses/total": 0.5973597168922424,
|
| 241 |
+
"ref_logps/chosen": -44.8856086730957,
|
| 242 |
+
"ref_logps/rejected": -60.589107513427734,
|
| 243 |
+
"rewards/accuracies": 0.6424999833106995,
|
| 244 |
+
"rewards/chosen": -0.15053769946098328,
|
| 245 |
+
"rewards/margins": 0.30677077174186707,
|
| 246 |
+
"rewards/rejected": -0.45730850100517273,
|
| 247 |
+
"step": 350
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.94,
|
| 251 |
+
"learning_rate": 3.819444444444444e-07,
|
| 252 |
+
"logps/chosen": -46.03995132446289,
|
| 253 |
+
"logps/rejected": -63.268062591552734,
|
| 254 |
+
"loss": 0.5952,
|
| 255 |
+
"losses/dpo": 0.5992311239242554,
|
| 256 |
+
"losses/sft": 1.2882429361343384,
|
| 257 |
+
"losses/total": 0.5992311239242554,
|
| 258 |
+
"ref_logps/chosen": -44.33562469482422,
|
| 259 |
+
"ref_logps/rejected": -58.439823150634766,
|
| 260 |
+
"rewards/accuracies": 0.6640000343322754,
|
| 261 |
+
"rewards/chosen": -0.17043215036392212,
|
| 262 |
+
"rewards/margins": 0.31239163875579834,
|
| 263 |
+
"rewards/rejected": -0.48282381892204285,
|
| 264 |
+
"step": 375
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 1.0,
|
| 268 |
+
"learning_rate": 3.703703703703703e-07,
|
| 269 |
+
"logps/chosen": -45.34758377075195,
|
| 270 |
+
"logps/rejected": -63.79766082763672,
|
| 271 |
+
"loss": 0.5798,
|
| 272 |
+
"losses/dpo": 0.573971152305603,
|
| 273 |
+
"losses/sft": 1.2241010665893555,
|
| 274 |
+
"losses/total": 0.573971152305603,
|
| 275 |
+
"ref_logps/chosen": -43.607730865478516,
|
| 276 |
+
"ref_logps/rejected": -58.310462951660156,
|
| 277 |
+
"rewards/accuracies": 0.6694999933242798,
|
| 278 |
+
"rewards/chosen": -0.17398566007614136,
|
| 279 |
+
"rewards/margins": 0.3747338056564331,
|
| 280 |
+
"rewards/rejected": -0.5487195253372192,
|
| 281 |
+
"step": 400
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 1.06,
|
| 285 |
+
"learning_rate": 3.587962962962963e-07,
|
| 286 |
+
"logps/chosen": -45.958282470703125,
|
| 287 |
+
"logps/rejected": -63.922576904296875,
|
| 288 |
+
"loss": 0.5783,
|
| 289 |
+
"losses/dpo": 0.5742554068565369,
|
| 290 |
+
"losses/sft": 1.2882492542266846,
|
| 291 |
+
"losses/total": 0.5742554068565369,
|
| 292 |
+
"ref_logps/chosen": -43.95150375366211,
|
| 293 |
+
"ref_logps/rejected": -58.02607345581055,
|
| 294 |
+
"rewards/accuracies": 0.6690000295639038,
|
| 295 |
+
"rewards/chosen": -0.20067782700061798,
|
| 296 |
+
"rewards/margins": 0.38897278904914856,
|
| 297 |
+
"rewards/rejected": -0.5896506309509277,
|
| 298 |
+
"step": 425
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 1.12,
|
| 302 |
+
"learning_rate": 3.472222222222222e-07,
|
| 303 |
+
"logps/chosen": -47.257179260253906,
|
| 304 |
+
"logps/rejected": -68.16796875,
|
| 305 |
+
"loss": 0.5616,
|
| 306 |
+
"losses/dpo": 0.5516221523284912,
|
| 307 |
+
"losses/sft": 1.2813299894332886,
|
| 308 |
+
"losses/total": 0.5516221523284912,
|
| 309 |
+
"ref_logps/chosen": -45.21623611450195,
|
| 310 |
+
"ref_logps/rejected": -61.444461822509766,
|
| 311 |
+
"rewards/accuracies": 0.6890000104904175,
|
| 312 |
+
"rewards/chosen": -0.20409366488456726,
|
| 313 |
+
"rewards/margins": 0.4682568609714508,
|
| 314 |
+
"rewards/rejected": -0.6723506450653076,
|
| 315 |
+
"step": 450
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 1.19,
|
| 319 |
+
"learning_rate": 3.3564814814814815e-07,
|
| 320 |
+
"logps/chosen": -46.08806610107422,
|
| 321 |
+
"logps/rejected": -65.21028137207031,
|
| 322 |
+
"loss": 0.5651,
|
| 323 |
+
"losses/dpo": 0.554095983505249,
|
| 324 |
+
"losses/sft": 1.3418306112289429,
|
| 325 |
+
"losses/total": 0.554095983505249,
|
| 326 |
+
"ref_logps/chosen": -43.922706604003906,
|
| 327 |
+
"ref_logps/rejected": -58.401084899902344,
|
| 328 |
+
"rewards/accuracies": 0.6819999814033508,
|
| 329 |
+
"rewards/chosen": -0.21653667092323303,
|
| 330 |
+
"rewards/margins": 0.4643844664096832,
|
| 331 |
+
"rewards/rejected": -0.680921196937561,
|
| 332 |
+
"step": 475
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 1.25,
|
| 336 |
+
"learning_rate": 3.2407407407407406e-07,
|
| 337 |
+
"logps/chosen": -46.008277893066406,
|
| 338 |
+
"logps/rejected": -66.00773620605469,
|
| 339 |
+
"loss": 0.5556,
|
| 340 |
+
"losses/dpo": 0.5685967803001404,
|
| 341 |
+
"losses/sft": 1.2652785778045654,
|
| 342 |
+
"losses/total": 0.5685967803001404,
|
| 343 |
+
"ref_logps/chosen": -43.85118103027344,
|
| 344 |
+
"ref_logps/rejected": -58.737857818603516,
|
| 345 |
+
"rewards/accuracies": 0.6944999694824219,
|
| 346 |
+
"rewards/chosen": -0.21570871770381927,
|
| 347 |
+
"rewards/margins": 0.5112798810005188,
|
| 348 |
+
"rewards/rejected": -0.7269885540008545,
|
| 349 |
+
"step": 500
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"epoch": 1.31,
|
| 353 |
+
"learning_rate": 3.1249999999999997e-07,
|
| 354 |
+
"logps/chosen": -45.792293548583984,
|
| 355 |
+
"logps/rejected": -68.81733703613281,
|
| 356 |
+
"loss": 0.5479,
|
| 357 |
+
"losses/dpo": 0.5653673410415649,
|
| 358 |
+
"losses/sft": 1.2111891508102417,
|
| 359 |
+
"losses/total": 0.5653673410415649,
|
| 360 |
+
"ref_logps/chosen": -43.507266998291016,
|
| 361 |
+
"ref_logps/rejected": -60.62786102294922,
|
| 362 |
+
"rewards/accuracies": 0.6984999775886536,
|
| 363 |
+
"rewards/chosen": -0.2285033017396927,
|
| 364 |
+
"rewards/margins": 0.5904435515403748,
|
| 365 |
+
"rewards/rejected": -0.8189470171928406,
|
| 366 |
+
"step": 525
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 1.38,
|
| 370 |
+
"learning_rate": 3.0092592592592594e-07,
|
| 371 |
+
"logps/chosen": -48.979248046875,
|
| 372 |
+
"logps/rejected": -70.8954849243164,
|
| 373 |
+
"loss": 0.5332,
|
| 374 |
+
"losses/dpo": 0.5159104466438293,
|
| 375 |
+
"losses/sft": 1.4047646522521973,
|
| 376 |
+
"losses/total": 0.5159104466438293,
|
| 377 |
+
"ref_logps/chosen": -46.33369064331055,
|
| 378 |
+
"ref_logps/rejected": -61.91999816894531,
|
| 379 |
+
"rewards/accuracies": 0.7049999833106995,
|
| 380 |
+
"rewards/chosen": -0.2645554840564728,
|
| 381 |
+
"rewards/margins": 0.6329929828643799,
|
| 382 |
+
"rewards/rejected": -0.897548496723175,
|
| 383 |
+
"step": 550
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 1.44,
|
| 387 |
+
"learning_rate": 2.8935185185185185e-07,
|
| 388 |
+
"logps/chosen": -47.573795318603516,
|
| 389 |
+
"logps/rejected": -71.3143081665039,
|
| 390 |
+
"loss": 0.5276,
|
| 391 |
+
"losses/dpo": 0.5273745656013489,
|
| 392 |
+
"losses/sft": 1.329495906829834,
|
| 393 |
+
"losses/total": 0.5273745656013489,
|
| 394 |
+
"ref_logps/chosen": -44.873233795166016,
|
| 395 |
+
"ref_logps/rejected": -61.62533187866211,
|
| 396 |
+
"rewards/accuracies": 0.6990000009536743,
|
| 397 |
+
"rewards/chosen": -0.27005600929260254,
|
| 398 |
+
"rewards/margins": 0.698841392993927,
|
| 399 |
+
"rewards/rejected": -0.9688975811004639,
|
| 400 |
+
"step": 575
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 1.5,
|
| 404 |
+
"learning_rate": 2.7777777777777776e-07,
|
| 405 |
+
"logps/chosen": -48.737342834472656,
|
| 406 |
+
"logps/rejected": -71.09253692626953,
|
| 407 |
+
"loss": 0.5234,
|
| 408 |
+
"losses/dpo": 0.5140664577484131,
|
| 409 |
+
"losses/sft": 1.3185381889343262,
|
| 410 |
+
"losses/total": 0.5140664577484131,
|
| 411 |
+
"ref_logps/chosen": -45.823787689208984,
|
| 412 |
+
"ref_logps/rejected": -60.519901275634766,
|
| 413 |
+
"rewards/accuracies": 0.7139999866485596,
|
| 414 |
+
"rewards/chosen": -0.2913552522659302,
|
| 415 |
+
"rewards/margins": 0.7659080624580383,
|
| 416 |
+
"rewards/rejected": -1.0572632551193237,
|
| 417 |
+
"step": 600
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 1.56,
|
| 421 |
+
"learning_rate": 2.662037037037037e-07,
|
| 422 |
+
"logps/chosen": -46.89459991455078,
|
| 423 |
+
"logps/rejected": -70.45913696289062,
|
| 424 |
+
"loss": 0.5194,
|
| 425 |
+
"losses/dpo": 0.5285363793373108,
|
| 426 |
+
"losses/sft": 1.3135132789611816,
|
| 427 |
+
"losses/total": 0.5285363793373108,
|
| 428 |
+
"ref_logps/chosen": -44.250526428222656,
|
| 429 |
+
"ref_logps/rejected": -59.6305046081543,
|
| 430 |
+
"rewards/accuracies": 0.7170000672340393,
|
| 431 |
+
"rewards/chosen": -0.2644067704677582,
|
| 432 |
+
"rewards/margins": 0.8184568881988525,
|
| 433 |
+
"rewards/rejected": -1.0828635692596436,
|
| 434 |
+
"step": 625
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 1.62,
|
| 438 |
+
"learning_rate": 2.5462962962962963e-07,
|
| 439 |
+
"logps/chosen": -48.4859619140625,
|
| 440 |
+
"logps/rejected": -71.74458312988281,
|
| 441 |
+
"loss": 0.522,
|
| 442 |
+
"losses/dpo": 0.5093265771865845,
|
| 443 |
+
"losses/sft": 1.342267632484436,
|
| 444 |
+
"losses/total": 0.5093265771865845,
|
| 445 |
+
"ref_logps/chosen": -45.3779182434082,
|
| 446 |
+
"ref_logps/rejected": -60.191993713378906,
|
| 447 |
+
"rewards/accuracies": 0.7124999761581421,
|
| 448 |
+
"rewards/chosen": -0.3108051121234894,
|
| 449 |
+
"rewards/margins": 0.8444538712501526,
|
| 450 |
+
"rewards/rejected": -1.1552588939666748,
|
| 451 |
+
"step": 650
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 1.69,
|
| 455 |
+
"learning_rate": 2.4305555555555555e-07,
|
| 456 |
+
"logps/chosen": -47.598331451416016,
|
| 457 |
+
"logps/rejected": -71.73502349853516,
|
| 458 |
+
"loss": 0.5051,
|
| 459 |
+
"losses/dpo": 0.5321754813194275,
|
| 460 |
+
"losses/sft": 1.288547158241272,
|
| 461 |
+
"losses/total": 0.5321754813194275,
|
| 462 |
+
"ref_logps/chosen": -44.57024383544922,
|
| 463 |
+
"ref_logps/rejected": -59.632904052734375,
|
| 464 |
+
"rewards/accuracies": 0.734499990940094,
|
| 465 |
+
"rewards/chosen": -0.3028090000152588,
|
| 466 |
+
"rewards/margins": 0.9074033498764038,
|
| 467 |
+
"rewards/rejected": -1.2102123498916626,
|
| 468 |
+
"step": 675
|
| 469 |
+
},
|
| 470 |
+
{
|
| 471 |
+
"epoch": 1.75,
|
| 472 |
+
"learning_rate": 2.3148148148148148e-07,
|
| 473 |
+
"logps/chosen": -48.9836540222168,
|
| 474 |
+
"logps/rejected": -75.99896240234375,
|
| 475 |
+
"loss": 0.5039,
|
| 476 |
+
"losses/dpo": 0.4898015558719635,
|
| 477 |
+
"losses/sft": 1.3348312377929688,
|
| 478 |
+
"losses/total": 0.4898015558719635,
|
| 479 |
+
"ref_logps/chosen": -45.258949279785156,
|
| 480 |
+
"ref_logps/rejected": -62.34097671508789,
|
| 481 |
+
"rewards/accuracies": 0.7209999561309814,
|
| 482 |
+
"rewards/chosen": -0.37246978282928467,
|
| 483 |
+
"rewards/margins": 0.993329644203186,
|
| 484 |
+
"rewards/rejected": -1.3657993078231812,
|
| 485 |
+
"step": 700
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 1.81,
|
| 489 |
+
"learning_rate": 2.199074074074074e-07,
|
| 490 |
+
"logps/chosen": -49.37858581542969,
|
| 491 |
+
"logps/rejected": -76.06786346435547,
|
| 492 |
+
"loss": 0.5037,
|
| 493 |
+
"losses/dpo": 0.4792703688144684,
|
| 494 |
+
"losses/sft": 1.3326802253723145,
|
| 495 |
+
"losses/total": 0.4792703688144684,
|
| 496 |
+
"ref_logps/chosen": -45.24223709106445,
|
| 497 |
+
"ref_logps/rejected": -61.8316764831543,
|
| 498 |
+
"rewards/accuracies": 0.7134999632835388,
|
| 499 |
+
"rewards/chosen": -0.4136350154876709,
|
| 500 |
+
"rewards/margins": 1.0099844932556152,
|
| 501 |
+
"rewards/rejected": -1.4236197471618652,
|
| 502 |
+
"step": 725
|
| 503 |
+
},
|
| 504 |
+
{
|
| 505 |
+
"epoch": 1.88,
|
| 506 |
+
"learning_rate": 2.0833333333333333e-07,
|
| 507 |
+
"logps/chosen": -48.57896041870117,
|
| 508 |
+
"logps/rejected": -71.33642578125,
|
| 509 |
+
"loss": 0.511,
|
| 510 |
+
"losses/dpo": 0.5458227396011353,
|
| 511 |
+
"losses/sft": 1.2892143726348877,
|
| 512 |
+
"losses/total": 0.5458227396011353,
|
| 513 |
+
"ref_logps/chosen": -44.61341857910156,
|
| 514 |
+
"ref_logps/rejected": -57.60921859741211,
|
| 515 |
+
"rewards/accuracies": 0.7119999527931213,
|
| 516 |
+
"rewards/chosen": -0.39655402302742004,
|
| 517 |
+
"rewards/margins": 0.9761665463447571,
|
| 518 |
+
"rewards/rejected": -1.3727205991744995,
|
| 519 |
+
"step": 750
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 1.94,
|
| 523 |
+
"learning_rate": 1.9675925925925927e-07,
|
| 524 |
+
"logps/chosen": -48.83073425292969,
|
| 525 |
+
"logps/rejected": -76.26995849609375,
|
| 526 |
+
"loss": 0.4896,
|
| 527 |
+
"losses/dpo": 0.4852687120437622,
|
| 528 |
+
"losses/sft": 1.3753533363342285,
|
| 529 |
+
"losses/total": 0.4852687120437622,
|
| 530 |
+
"ref_logps/chosen": -44.98310089111328,
|
| 531 |
+
"ref_logps/rejected": -61.261417388916016,
|
| 532 |
+
"rewards/accuracies": 0.7349998950958252,
|
| 533 |
+
"rewards/chosen": -0.38476306200027466,
|
| 534 |
+
"rewards/margins": 1.1160913705825806,
|
| 535 |
+
"rewards/rejected": -1.5008544921875,
|
| 536 |
+
"step": 775
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 2.0,
|
| 540 |
+
"learning_rate": 1.8518518518518516e-07,
|
| 541 |
+
"logps/chosen": -49.678104400634766,
|
| 542 |
+
"logps/rejected": -76.78030395507812,
|
| 543 |
+
"loss": 0.5058,
|
| 544 |
+
"losses/dpo": 0.4918399751186371,
|
| 545 |
+
"losses/sft": 1.373268485069275,
|
| 546 |
+
"losses/total": 0.4918399751186371,
|
| 547 |
+
"ref_logps/chosen": -45.21029281616211,
|
| 548 |
+
"ref_logps/rejected": -61.68964767456055,
|
| 549 |
+
"rewards/accuracies": 0.7149999737739563,
|
| 550 |
+
"rewards/chosen": -0.44678181409835815,
|
| 551 |
+
"rewards/margins": 1.0622824430465698,
|
| 552 |
+
"rewards/rejected": -1.5090643167495728,
|
| 553 |
+
"step": 800
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 2.06,
|
| 557 |
+
"learning_rate": 1.736111111111111e-07,
|
| 558 |
+
"logps/chosen": -49.76744079589844,
|
| 559 |
+
"logps/rejected": -77.67528533935547,
|
| 560 |
+
"loss": 0.4877,
|
| 561 |
+
"losses/dpo": 0.47753456234931946,
|
| 562 |
+
"losses/sft": 1.3492231369018555,
|
| 563 |
+
"losses/total": 0.47753456234931946,
|
| 564 |
+
"ref_logps/chosen": -45.20912170410156,
|
| 565 |
+
"ref_logps/rejected": -61.64081954956055,
|
| 566 |
+
"rewards/accuracies": 0.7359998822212219,
|
| 567 |
+
"rewards/chosen": -0.45583218336105347,
|
| 568 |
+
"rewards/margins": 1.1476140022277832,
|
| 569 |
+
"rewards/rejected": -1.6034462451934814,
|
| 570 |
+
"step": 825
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 2.12,
|
| 574 |
+
"learning_rate": 1.6203703703703703e-07,
|
| 575 |
+
"logps/chosen": -48.476280212402344,
|
| 576 |
+
"logps/rejected": -74.0830078125,
|
| 577 |
+
"loss": 0.4932,
|
| 578 |
+
"losses/dpo": 0.4669117331504822,
|
| 579 |
+
"losses/sft": 1.3778959512710571,
|
| 580 |
+
"losses/total": 0.4669117331504822,
|
| 581 |
+
"ref_logps/chosen": -44.10759735107422,
|
| 582 |
+
"ref_logps/rejected": -58.72689437866211,
|
| 583 |
+
"rewards/accuracies": 0.7190000414848328,
|
| 584 |
+
"rewards/chosen": -0.43686822056770325,
|
| 585 |
+
"rewards/margins": 1.098743200302124,
|
| 586 |
+
"rewards/rejected": -1.5356113910675049,
|
| 587 |
+
"step": 850
|
| 588 |
+
},
|
| 589 |
+
{
|
| 590 |
+
"epoch": 2.19,
|
| 591 |
+
"learning_rate": 1.5046296296296297e-07,
|
| 592 |
+
"logps/chosen": -49.38998031616211,
|
| 593 |
+
"logps/rejected": -76.47419738769531,
|
| 594 |
+
"loss": 0.5013,
|
| 595 |
+
"losses/dpo": 0.4948745667934418,
|
| 596 |
+
"losses/sft": 1.3688327074050903,
|
| 597 |
+
"losses/total": 0.4948745667934418,
|
| 598 |
+
"ref_logps/chosen": -44.37105941772461,
|
| 599 |
+
"ref_logps/rejected": -60.47099304199219,
|
| 600 |
+
"rewards/accuracies": 0.7259998917579651,
|
| 601 |
+
"rewards/chosen": -0.5018922686576843,
|
| 602 |
+
"rewards/margins": 1.0984277725219727,
|
| 603 |
+
"rewards/rejected": -1.6003201007843018,
|
| 604 |
+
"step": 875
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 2.25,
|
| 608 |
+
"learning_rate": 1.3888888888888888e-07,
|
| 609 |
+
"logps/chosen": -50.32158279418945,
|
| 610 |
+
"logps/rejected": -77.23485565185547,
|
| 611 |
+
"loss": 0.4968,
|
| 612 |
+
"losses/dpo": 0.5039748549461365,
|
| 613 |
+
"losses/sft": 1.3766621351242065,
|
| 614 |
+
"losses/total": 0.5039748549461365,
|
| 615 |
+
"ref_logps/chosen": -45.48880386352539,
|
| 616 |
+
"ref_logps/rejected": -60.92485809326172,
|
| 617 |
+
"rewards/accuracies": 0.7320000529289246,
|
| 618 |
+
"rewards/chosen": -0.48327693343162537,
|
| 619 |
+
"rewards/margins": 1.1477227210998535,
|
| 620 |
+
"rewards/rejected": -1.6309998035430908,
|
| 621 |
+
"step": 900
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 2.31,
|
| 625 |
+
"learning_rate": 1.2731481481481482e-07,
|
| 626 |
+
"logps/chosen": -49.106754302978516,
|
| 627 |
+
"logps/rejected": -75.4212417602539,
|
| 628 |
+
"loss": 0.5004,
|
| 629 |
+
"losses/dpo": 0.465129017829895,
|
| 630 |
+
"losses/sft": 1.396937370300293,
|
| 631 |
+
"losses/total": 0.465129017829895,
|
| 632 |
+
"ref_logps/chosen": -44.43598556518555,
|
| 633 |
+
"ref_logps/rejected": -59.51651382446289,
|
| 634 |
+
"rewards/accuracies": 0.7200000286102295,
|
| 635 |
+
"rewards/chosen": -0.4670768678188324,
|
| 636 |
+
"rewards/margins": 1.123395562171936,
|
| 637 |
+
"rewards/rejected": -1.5904724597930908,
|
| 638 |
+
"step": 925
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 2.38,
|
| 642 |
+
"learning_rate": 1.1574074074074074e-07,
|
| 643 |
+
"logps/chosen": -49.315948486328125,
|
| 644 |
+
"logps/rejected": -75.77275848388672,
|
| 645 |
+
"loss": 0.4952,
|
| 646 |
+
"losses/dpo": 0.48752278089523315,
|
| 647 |
+
"losses/sft": 1.383622646331787,
|
| 648 |
+
"losses/total": 0.48752278089523315,
|
| 649 |
+
"ref_logps/chosen": -44.68310546875,
|
| 650 |
+
"ref_logps/rejected": -59.506874084472656,
|
| 651 |
+
"rewards/accuracies": 0.7290000915527344,
|
| 652 |
+
"rewards/chosen": -0.4632847309112549,
|
| 653 |
+
"rewards/margins": 1.1633038520812988,
|
| 654 |
+
"rewards/rejected": -1.6265885829925537,
|
| 655 |
+
"step": 950
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 2.44,
|
| 659 |
+
"learning_rate": 1.0416666666666667e-07,
|
| 660 |
+
"logps/chosen": -49.365966796875,
|
| 661 |
+
"logps/rejected": -76.18710327148438,
|
| 662 |
+
"loss": 0.4945,
|
| 663 |
+
"losses/dpo": 0.5000871419906616,
|
| 664 |
+
"losses/sft": 1.345529317855835,
|
| 665 |
+
"losses/total": 0.5000871419906616,
|
| 666 |
+
"ref_logps/chosen": -44.74851989746094,
|
| 667 |
+
"ref_logps/rejected": -60.123268127441406,
|
| 668 |
+
"rewards/accuracies": 0.7280000448226929,
|
| 669 |
+
"rewards/chosen": -0.46174487471580505,
|
| 670 |
+
"rewards/margins": 1.1446377038955688,
|
| 671 |
+
"rewards/rejected": -1.6063826084136963,
|
| 672 |
+
"step": 975
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 2.5,
|
| 676 |
+
"learning_rate": 9.259259259259258e-08,
|
| 677 |
+
"logps/chosen": -49.790008544921875,
|
| 678 |
+
"logps/rejected": -76.8461685180664,
|
| 679 |
+
"loss": 0.4891,
|
| 680 |
+
"losses/dpo": 0.5052517056465149,
|
| 681 |
+
"losses/sft": 1.3480595350265503,
|
| 682 |
+
"losses/total": 0.5052517056465149,
|
| 683 |
+
"ref_logps/chosen": -44.7381477355957,
|
| 684 |
+
"ref_logps/rejected": -60.120697021484375,
|
| 685 |
+
"rewards/accuracies": 0.7300000786781311,
|
| 686 |
+
"rewards/chosen": -0.505186140537262,
|
| 687 |
+
"rewards/margins": 1.167360782623291,
|
| 688 |
+
"rewards/rejected": -1.6725467443466187,
|
| 689 |
+
"step": 1000
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 2.56,
|
| 693 |
+
"learning_rate": 8.101851851851852e-08,
|
| 694 |
+
"logps/chosen": -48.60505676269531,
|
| 695 |
+
"logps/rejected": -75.21623992919922,
|
| 696 |
+
"loss": 0.4956,
|
| 697 |
+
"losses/dpo": 0.4796887934207916,
|
| 698 |
+
"losses/sft": 1.3435637950897217,
|
| 699 |
+
"losses/total": 0.4796887934207916,
|
| 700 |
+
"ref_logps/chosen": -43.89369583129883,
|
| 701 |
+
"ref_logps/rejected": -59.03510665893555,
|
| 702 |
+
"rewards/accuracies": 0.718000054359436,
|
| 703 |
+
"rewards/chosen": -0.47113633155822754,
|
| 704 |
+
"rewards/margins": 1.146977186203003,
|
| 705 |
+
"rewards/rejected": -1.618113398551941,
|
| 706 |
+
"step": 1025
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 2.62,
|
| 710 |
+
"learning_rate": 6.944444444444444e-08,
|
| 711 |
+
"logps/chosen": -49.14812469482422,
|
| 712 |
+
"logps/rejected": -77.22814178466797,
|
| 713 |
+
"loss": 0.4816,
|
| 714 |
+
"losses/dpo": 0.49271050095558167,
|
| 715 |
+
"losses/sft": 1.3505686521530151,
|
| 716 |
+
"losses/total": 0.49271050095558167,
|
| 717 |
+
"ref_logps/chosen": -44.201717376708984,
|
| 718 |
+
"ref_logps/rejected": -60.2335205078125,
|
| 719 |
+
"rewards/accuracies": 0.7355000376701355,
|
| 720 |
+
"rewards/chosen": -0.4946403503417969,
|
| 721 |
+
"rewards/margins": 1.2048207521438599,
|
| 722 |
+
"rewards/rejected": -1.6994611024856567,
|
| 723 |
+
"step": 1050
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 2.69,
|
| 727 |
+
"learning_rate": 5.787037037037037e-08,
|
| 728 |
+
"logps/chosen": -50.257972717285156,
|
| 729 |
+
"logps/rejected": -79.50613403320312,
|
| 730 |
+
"loss": 0.4888,
|
| 731 |
+
"losses/dpo": 0.4835771918296814,
|
| 732 |
+
"losses/sft": 1.399458408355713,
|
| 733 |
+
"losses/total": 0.4835771918296814,
|
| 734 |
+
"ref_logps/chosen": -44.78007888793945,
|
| 735 |
+
"ref_logps/rejected": -61.73722839355469,
|
| 736 |
+
"rewards/accuracies": 0.7374999523162842,
|
| 737 |
+
"rewards/chosen": -0.5477903485298157,
|
| 738 |
+
"rewards/margins": 1.2291010618209839,
|
| 739 |
+
"rewards/rejected": -1.7768914699554443,
|
| 740 |
+
"step": 1075
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 2.75,
|
| 744 |
+
"learning_rate": 4.629629629629629e-08,
|
| 745 |
+
"logps/chosen": -50.44590759277344,
|
| 746 |
+
"logps/rejected": -80.34747314453125,
|
| 747 |
+
"loss": 0.4749,
|
| 748 |
+
"losses/dpo": 0.4630358815193176,
|
| 749 |
+
"losses/sft": 1.3918135166168213,
|
| 750 |
+
"losses/total": 0.4630358815193176,
|
| 751 |
+
"ref_logps/chosen": -45.468910217285156,
|
| 752 |
+
"ref_logps/rejected": -62.39448547363281,
|
| 753 |
+
"rewards/accuracies": 0.7430000305175781,
|
| 754 |
+
"rewards/chosen": -0.497699499130249,
|
| 755 |
+
"rewards/margins": 1.297598123550415,
|
| 756 |
+
"rewards/rejected": -1.7952975034713745,
|
| 757 |
+
"step": 1100
|
| 758 |
+
},
|
| 759 |
+
{
|
| 760 |
+
"epoch": 2.81,
|
| 761 |
+
"learning_rate": 3.472222222222222e-08,
|
| 762 |
+
"logps/chosen": -50.564212799072266,
|
| 763 |
+
"logps/rejected": -77.6652603149414,
|
| 764 |
+
"loss": 0.4844,
|
| 765 |
+
"losses/dpo": 0.4967005252838135,
|
| 766 |
+
"losses/sft": 1.3921661376953125,
|
| 767 |
+
"losses/total": 0.4967005252838135,
|
| 768 |
+
"ref_logps/chosen": -45.257930755615234,
|
| 769 |
+
"ref_logps/rejected": -59.98822784423828,
|
| 770 |
+
"rewards/accuracies": 0.7325000166893005,
|
| 771 |
+
"rewards/chosen": -0.5306286811828613,
|
| 772 |
+
"rewards/margins": 1.237074613571167,
|
| 773 |
+
"rewards/rejected": -1.7677034139633179,
|
| 774 |
+
"step": 1125
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 2.88,
|
| 778 |
+
"learning_rate": 2.3148148148148144e-08,
|
| 779 |
+
"logps/chosen": -49.5255126953125,
|
| 780 |
+
"logps/rejected": -78.27824401855469,
|
| 781 |
+
"loss": 0.4747,
|
| 782 |
+
"losses/dpo": 0.48569077253341675,
|
| 783 |
+
"losses/sft": 1.3604434728622437,
|
| 784 |
+
"losses/total": 0.48569077253341675,
|
| 785 |
+
"ref_logps/chosen": -44.655574798583984,
|
| 786 |
+
"ref_logps/rejected": -60.496063232421875,
|
| 787 |
+
"rewards/accuracies": 0.7485000491142273,
|
| 788 |
+
"rewards/chosen": -0.48699355125427246,
|
| 789 |
+
"rewards/margins": 1.2912240028381348,
|
| 790 |
+
"rewards/rejected": -1.7782177925109863,
|
| 791 |
+
"step": 1150
|
| 792 |
+
},
|
| 793 |
+
{
|
| 794 |
+
"epoch": 2.94,
|
| 795 |
+
"learning_rate": 1.1574074074074072e-08,
|
| 796 |
+
"logps/chosen": -50.011077880859375,
|
| 797 |
+
"logps/rejected": -75.67598724365234,
|
| 798 |
+
"loss": 0.4876,
|
| 799 |
+
"losses/dpo": 0.5032810568809509,
|
| 800 |
+
"losses/sft": 1.366944432258606,
|
| 801 |
+
"losses/total": 0.5032810568809509,
|
| 802 |
+
"ref_logps/chosen": -44.80411148071289,
|
| 803 |
+
"ref_logps/rejected": -58.415794372558594,
|
| 804 |
+
"rewards/accuracies": 0.7414999604225159,
|
| 805 |
+
"rewards/chosen": -0.5206969976425171,
|
| 806 |
+
"rewards/margins": 1.2053215503692627,
|
| 807 |
+
"rewards/rejected": -1.7260186672210693,
|
| 808 |
+
"step": 1175
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 3.0,
|
| 812 |
+
"learning_rate": 0.0,
|
| 813 |
+
"logps/chosen": -51.40888595581055,
|
| 814 |
+
"logps/rejected": -80.48424530029297,
|
| 815 |
+
"loss": 0.469,
|
| 816 |
+
"losses/dpo": 0.4639938771724701,
|
| 817 |
+
"losses/sft": 1.3975021839141846,
|
| 818 |
+
"losses/total": 0.4639938771724701,
|
| 819 |
+
"ref_logps/chosen": -46.229427337646484,
|
| 820 |
+
"ref_logps/rejected": -62.2370491027832,
|
| 821 |
+
"rewards/accuracies": 0.7464999556541443,
|
| 822 |
+
"rewards/chosen": -0.5179460644721985,
|
| 823 |
+
"rewards/margins": 1.3067736625671387,
|
| 824 |
+
"rewards/rejected": -1.8247196674346924,
|
| 825 |
+
"step": 1200
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 3.0,
|
| 829 |
+
"step": 1200,
|
| 830 |
+
"total_flos": 0.0,
|
| 831 |
+
"train_loss": 0.5544587286313375,
|
| 832 |
+
"train_runtime": 33858.4559,
|
| 833 |
+
"train_samples_per_second": 2.835,
|
| 834 |
+
"train_steps_per_second": 0.035
|
| 835 |
+
}
|
| 836 |
+
],
|
| 837 |
+
"logging_steps": 25,
|
| 838 |
+
"max_steps": 1200,
|
| 839 |
+
"num_input_tokens_seen": 0,
|
| 840 |
+
"num_train_epochs": 3,
|
| 841 |
+
"save_steps": 24000,
|
| 842 |
+
"total_flos": 0.0,
|
| 843 |
+
"train_batch_size": 4,
|
| 844 |
+
"trial_name": null,
|
| 845 |
+
"trial_params": null
|
| 846 |
+
}
|