Upload folder using huggingface_hub
Browse files- README.md +9 -0
- adapter_config.json +29 -0
- adapter_model.bin +3 -0
- config.json +56 -0
- non_lora_trainables.bin +3 -0
- trainer_state.json +830 -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 +1,29 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"auto_mapping": null,
|
| 3 |
+
"base_model_name_or_path": "DAMO-NLP-SG/VideoLLaMA3-7B",
|
| 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 |
+
"gate_proj",
|
| 18 |
+
"out_proj",
|
| 19 |
+
"down_proj",
|
| 20 |
+
"q_proj",
|
| 21 |
+
"fc2",
|
| 22 |
+
"v_proj",
|
| 23 |
+
"o_proj",
|
| 24 |
+
"up_proj",
|
| 25 |
+
"k_proj",
|
| 26 |
+
"fc1"
|
| 27 |
+
],
|
| 28 |
+
"task_type": "CAUSAL_LM"
|
| 29 |
+
}
|
adapter_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1be9e7b0fba83d5e831d60538adce9138fc2aa347d6ff2ca540ae992a8f10f61
|
| 3 |
+
size 785303482
|
config.json
ADDED
|
@@ -0,0 +1,56 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"X": [
|
| 3 |
+
"Image",
|
| 4 |
+
"Video"
|
| 5 |
+
],
|
| 6 |
+
"_attn_implementation_autoset": true,
|
| 7 |
+
"_name_or_path": "DAMO-NLP-SG/VideoLLaMA3-7B",
|
| 8 |
+
"architectures": [
|
| 9 |
+
"Videollama3Qwen2ForCausalLM"
|
| 10 |
+
],
|
| 11 |
+
"attention_dropout": 0.0,
|
| 12 |
+
"auto_map": {
|
| 13 |
+
"AutoConfig": "DAMO-NLP-SG/VideoLLaMA3-7B--configuration_videollama3.Videollama3Qwen2Config",
|
| 14 |
+
"AutoModelForCausalLM": "DAMO-NLP-SG/VideoLLaMA3-7B--modeling_videollama3.Videollama3Qwen2ForCausalLM"
|
| 15 |
+
},
|
| 16 |
+
"bos_token_id": 151643,
|
| 17 |
+
"eos_token_id": 151645,
|
| 18 |
+
"freeze_mm_mlp_adapter": true,
|
| 19 |
+
"hidden_act": "silu",
|
| 20 |
+
"hidden_size": 3584,
|
| 21 |
+
"image_aspect_ratio": "pad",
|
| 22 |
+
"image_grid_pinpoints": null,
|
| 23 |
+
"image_token_index": 151665,
|
| 24 |
+
"initializer_range": 0.02,
|
| 25 |
+
"intermediate_size": 18944,
|
| 26 |
+
"max_position_embeddings": 32768,
|
| 27 |
+
"max_window_layers": 28,
|
| 28 |
+
"mm_projector_type": "mlp2x_gelu",
|
| 29 |
+
"mm_use_x_patch_token": false,
|
| 30 |
+
"mm_use_x_start_end": false,
|
| 31 |
+
"model_type": "videollama3_qwen2",
|
| 32 |
+
"num_attention_heads": 28,
|
| 33 |
+
"num_hidden_layers": 28,
|
| 34 |
+
"num_key_value_heads": 4,
|
| 35 |
+
"rms_norm_eps": 1e-06,
|
| 36 |
+
"rope_scaling": null,
|
| 37 |
+
"rope_theta": 1000000.0,
|
| 38 |
+
"sliding_window": null,
|
| 39 |
+
"tie_word_embeddings": false,
|
| 40 |
+
"torch_dtype": "bfloat16",
|
| 41 |
+
"transformers_version": "4.46.3",
|
| 42 |
+
"tune_mm_mlp_adapter": false,
|
| 43 |
+
"use_cache": true,
|
| 44 |
+
"use_sliding_window": false,
|
| 45 |
+
"use_token_compression": false,
|
| 46 |
+
"vision_encoder": null,
|
| 47 |
+
"vision_encoder_config": {
|
| 48 |
+
"hidden_size": 1152,
|
| 49 |
+
"intermediate_size": 4304,
|
| 50 |
+
"model_type": "videollama3_vision_encoder",
|
| 51 |
+
"num_attention_heads": 16,
|
| 52 |
+
"num_hidden_layers": 27,
|
| 53 |
+
"patch_size": 14
|
| 54 |
+
},
|
| 55 |
+
"vocab_size": 152064
|
| 56 |
+
}
|
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 +1,830 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.9962453066332917,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 1197,
|
| 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": -71.84510803222656,
|
| 15 |
+
"logps/rejected": -68.73027038574219,
|
| 16 |
+
"loss": 0.7215,
|
| 17 |
+
"losses/dpo": 0.7109084129333496,
|
| 18 |
+
"losses/sft": 1.546567678451538,
|
| 19 |
+
"losses/total": 0.7109084129333496,
|
| 20 |
+
"ref_logps/chosen": -71.75371551513672,
|
| 21 |
+
"ref_logps/rejected": -68.71001434326172,
|
| 22 |
+
"rewards/accuracies": 0.5049999952316284,
|
| 23 |
+
"rewards/chosen": -0.009138082154095173,
|
| 24 |
+
"rewards/margins": -0.0071117933839559555,
|
| 25 |
+
"rewards/rejected": -0.0020262906327843666,
|
| 26 |
+
"step": 25
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.13,
|
| 30 |
+
"learning_rate": 2.0833333333333333e-07,
|
| 31 |
+
"logps/chosen": -69.88240814208984,
|
| 32 |
+
"logps/rejected": -69.54386901855469,
|
| 33 |
+
"loss": 0.7126,
|
| 34 |
+
"losses/dpo": 0.6903221607208252,
|
| 35 |
+
"losses/sft": 1.5194830894470215,
|
| 36 |
+
"losses/total": 0.6903221607208252,
|
| 37 |
+
"ref_logps/chosen": -69.94444274902344,
|
| 38 |
+
"ref_logps/rejected": -69.52371978759766,
|
| 39 |
+
"rewards/accuracies": 0.4970000088214874,
|
| 40 |
+
"rewards/chosen": 0.006204119883477688,
|
| 41 |
+
"rewards/margins": 0.008218951523303986,
|
| 42 |
+
"rewards/rejected": -0.002014830242842436,
|
| 43 |
+
"step": 50
|
| 44 |
+
},
|
| 45 |
+
{
|
| 46 |
+
"epoch": 0.19,
|
| 47 |
+
"learning_rate": 3.1249999999999997e-07,
|
| 48 |
+
"logps/chosen": -73.9260482788086,
|
| 49 |
+
"logps/rejected": -71.71464538574219,
|
| 50 |
+
"loss": 0.7231,
|
| 51 |
+
"losses/dpo": 0.7194635272026062,
|
| 52 |
+
"losses/sft": 1.5661953687667847,
|
| 53 |
+
"losses/total": 0.7194635272026062,
|
| 54 |
+
"ref_logps/chosen": -73.91621398925781,
|
| 55 |
+
"ref_logps/rejected": -71.82748413085938,
|
| 56 |
+
"rewards/accuracies": 0.5029999613761902,
|
| 57 |
+
"rewards/chosen": -0.0009833112126216292,
|
| 58 |
+
"rewards/margins": -0.01226747315376997,
|
| 59 |
+
"rewards/rejected": 0.011284159496426582,
|
| 60 |
+
"step": 75
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 0.25,
|
| 64 |
+
"learning_rate": 4.1666666666666667e-07,
|
| 65 |
+
"logps/chosen": -68.9831314086914,
|
| 66 |
+
"logps/rejected": -66.86358642578125,
|
| 67 |
+
"loss": 0.7192,
|
| 68 |
+
"losses/dpo": 0.7422583103179932,
|
| 69 |
+
"losses/sft": 1.5391790866851807,
|
| 70 |
+
"losses/total": 0.7422583103179932,
|
| 71 |
+
"ref_logps/chosen": -68.984130859375,
|
| 72 |
+
"ref_logps/rejected": -66.90238189697266,
|
| 73 |
+
"rewards/accuracies": 0.4930000305175781,
|
| 74 |
+
"rewards/chosen": 0.00010019920591730624,
|
| 75 |
+
"rewards/margins": -0.00377923552878201,
|
| 76 |
+
"rewards/rejected": 0.003879436058923602,
|
| 77 |
+
"step": 100
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 0.31,
|
| 81 |
+
"learning_rate": 4.976787372330547e-07,
|
| 82 |
+
"logps/chosen": -73.55087280273438,
|
| 83 |
+
"logps/rejected": -71.61868286132812,
|
| 84 |
+
"loss": 0.7077,
|
| 85 |
+
"losses/dpo": 0.7084662914276123,
|
| 86 |
+
"losses/sft": 1.5277128219604492,
|
| 87 |
+
"losses/total": 0.7084662914276123,
|
| 88 |
+
"ref_logps/chosen": -73.66172790527344,
|
| 89 |
+
"ref_logps/rejected": -71.5505599975586,
|
| 90 |
+
"rewards/accuracies": 0.5219999551773071,
|
| 91 |
+
"rewards/chosen": 0.011085684411227703,
|
| 92 |
+
"rewards/margins": 0.017898183315992355,
|
| 93 |
+
"rewards/rejected": -0.00681249750778079,
|
| 94 |
+
"step": 125
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.38,
|
| 98 |
+
"learning_rate": 4.860724233983287e-07,
|
| 99 |
+
"logps/chosen": -69.85209655761719,
|
| 100 |
+
"logps/rejected": -69.75456237792969,
|
| 101 |
+
"loss": 0.7016,
|
| 102 |
+
"losses/dpo": 0.6934231519699097,
|
| 103 |
+
"losses/sft": 1.5373976230621338,
|
| 104 |
+
"losses/total": 0.6934231519699097,
|
| 105 |
+
"ref_logps/chosen": -70.19540405273438,
|
| 106 |
+
"ref_logps/rejected": -69.83726501464844,
|
| 107 |
+
"rewards/accuracies": 0.5325000286102295,
|
| 108 |
+
"rewards/chosen": 0.0343310683965683,
|
| 109 |
+
"rewards/margins": 0.026060676202178,
|
| 110 |
+
"rewards/rejected": 0.008270387537777424,
|
| 111 |
+
"step": 150
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.44,
|
| 115 |
+
"learning_rate": 4.7446610956360255e-07,
|
| 116 |
+
"logps/chosen": -69.9237289428711,
|
| 117 |
+
"logps/rejected": -69.56204223632812,
|
| 118 |
+
"loss": 0.6826,
|
| 119 |
+
"losses/dpo": 0.6780227422714233,
|
| 120 |
+
"losses/sft": 1.5951652526855469,
|
| 121 |
+
"losses/total": 0.6780227422714233,
|
| 122 |
+
"ref_logps/chosen": -70.6123046875,
|
| 123 |
+
"ref_logps/rejected": -69.56880187988281,
|
| 124 |
+
"rewards/accuracies": 0.5539999604225159,
|
| 125 |
+
"rewards/chosen": 0.06885794550180435,
|
| 126 |
+
"rewards/margins": 0.0681825652718544,
|
| 127 |
+
"rewards/rejected": 0.0006753735360689461,
|
| 128 |
+
"step": 175
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.5,
|
| 132 |
+
"learning_rate": 4.628597957288765e-07,
|
| 133 |
+
"logps/chosen": -71.02560424804688,
|
| 134 |
+
"logps/rejected": -69.64324951171875,
|
| 135 |
+
"loss": 0.679,
|
| 136 |
+
"losses/dpo": 0.6822719573974609,
|
| 137 |
+
"losses/sft": 1.5291378498077393,
|
| 138 |
+
"losses/total": 0.6822719573974609,
|
| 139 |
+
"ref_logps/chosen": -71.80342102050781,
|
| 140 |
+
"ref_logps/rejected": -69.58236694335938,
|
| 141 |
+
"rewards/accuracies": 0.5559999942779541,
|
| 142 |
+
"rewards/chosen": 0.07778114080429077,
|
| 143 |
+
"rewards/margins": 0.08386911451816559,
|
| 144 |
+
"rewards/rejected": -0.006087968125939369,
|
| 145 |
+
"step": 200
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 0.56,
|
| 149 |
+
"learning_rate": 4.512534818941504e-07,
|
| 150 |
+
"logps/chosen": -67.88444519042969,
|
| 151 |
+
"logps/rejected": -68.10497283935547,
|
| 152 |
+
"loss": 0.679,
|
| 153 |
+
"losses/dpo": 0.6802477240562439,
|
| 154 |
+
"losses/sft": 1.5126712322235107,
|
| 155 |
+
"losses/total": 0.6802477240562439,
|
| 156 |
+
"ref_logps/chosen": -68.86332702636719,
|
| 157 |
+
"ref_logps/rejected": -68.26943969726562,
|
| 158 |
+
"rewards/accuracies": 0.5540000200271606,
|
| 159 |
+
"rewards/chosen": 0.0978882685303688,
|
| 160 |
+
"rewards/margins": 0.08144120872020721,
|
| 161 |
+
"rewards/rejected": 0.01644706539809704,
|
| 162 |
+
"step": 225
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 0.63,
|
| 166 |
+
"learning_rate": 4.3964716805942433e-07,
|
| 167 |
+
"logps/chosen": -69.61144256591797,
|
| 168 |
+
"logps/rejected": -71.70446014404297,
|
| 169 |
+
"loss": 0.661,
|
| 170 |
+
"losses/dpo": 0.6608595848083496,
|
| 171 |
+
"losses/sft": 1.5144795179367065,
|
| 172 |
+
"losses/total": 0.6608595848083496,
|
| 173 |
+
"ref_logps/chosen": -70.85598754882812,
|
| 174 |
+
"ref_logps/rejected": -71.67627716064453,
|
| 175 |
+
"rewards/accuracies": 0.5940000414848328,
|
| 176 |
+
"rewards/chosen": 0.12445437163114548,
|
| 177 |
+
"rewards/margins": 0.12727266550064087,
|
| 178 |
+
"rewards/rejected": -0.0028182892128825188,
|
| 179 |
+
"step": 250
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 0.69,
|
| 183 |
+
"learning_rate": 4.280408542246982e-07,
|
| 184 |
+
"logps/chosen": -69.20745849609375,
|
| 185 |
+
"logps/rejected": -68.46233367919922,
|
| 186 |
+
"loss": 0.6577,
|
| 187 |
+
"losses/dpo": 0.669312059879303,
|
| 188 |
+
"losses/sft": 1.5866602659225464,
|
| 189 |
+
"losses/total": 0.669312059879303,
|
| 190 |
+
"ref_logps/chosen": -70.55343627929688,
|
| 191 |
+
"ref_logps/rejected": -68.40556335449219,
|
| 192 |
+
"rewards/accuracies": 0.5860000252723694,
|
| 193 |
+
"rewards/chosen": 0.13459768891334534,
|
| 194 |
+
"rewards/margins": 0.14027482271194458,
|
| 195 |
+
"rewards/rejected": -0.005677163600921631,
|
| 196 |
+
"step": 275
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 0.75,
|
| 200 |
+
"learning_rate": 4.164345403899721e-07,
|
| 201 |
+
"logps/chosen": -70.11485290527344,
|
| 202 |
+
"logps/rejected": -68.45763397216797,
|
| 203 |
+
"loss": 0.649,
|
| 204 |
+
"losses/dpo": 0.6586881875991821,
|
| 205 |
+
"losses/sft": 1.5220152139663696,
|
| 206 |
+
"losses/total": 0.6586881875991821,
|
| 207 |
+
"ref_logps/chosen": -71.6851806640625,
|
| 208 |
+
"ref_logps/rejected": -68.39569091796875,
|
| 209 |
+
"rewards/accuracies": 0.5975000262260437,
|
| 210 |
+
"rewards/chosen": 0.15703237056732178,
|
| 211 |
+
"rewards/margins": 0.16322720050811768,
|
| 212 |
+
"rewards/rejected": -0.006194834131747484,
|
| 213 |
+
"step": 300
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.81,
|
| 217 |
+
"learning_rate": 4.04828226555246e-07,
|
| 218 |
+
"logps/chosen": -69.53617095947266,
|
| 219 |
+
"logps/rejected": -68.70401763916016,
|
| 220 |
+
"loss": 0.6401,
|
| 221 |
+
"losses/dpo": 0.6386440396308899,
|
| 222 |
+
"losses/sft": 1.4834158420562744,
|
| 223 |
+
"losses/total": 0.6386440396308899,
|
| 224 |
+
"ref_logps/chosen": -71.28308868408203,
|
| 225 |
+
"ref_logps/rejected": -68.5333480834961,
|
| 226 |
+
"rewards/accuracies": 0.6014999151229858,
|
| 227 |
+
"rewards/chosen": 0.17469124495983124,
|
| 228 |
+
"rewards/margins": 0.19175761938095093,
|
| 229 |
+
"rewards/rejected": -0.017066391184926033,
|
| 230 |
+
"step": 325
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.88,
|
| 234 |
+
"learning_rate": 3.9322191272051997e-07,
|
| 235 |
+
"logps/chosen": -69.60104370117188,
|
| 236 |
+
"logps/rejected": -70.08628845214844,
|
| 237 |
+
"loss": 0.6349,
|
| 238 |
+
"losses/dpo": 0.649332582950592,
|
| 239 |
+
"losses/sft": 1.4883217811584473,
|
| 240 |
+
"losses/total": 0.649332582950592,
|
| 241 |
+
"ref_logps/chosen": -71.48506164550781,
|
| 242 |
+
"ref_logps/rejected": -69.83267211914062,
|
| 243 |
+
"rewards/accuracies": 0.6045000553131104,
|
| 244 |
+
"rewards/chosen": 0.18840213119983673,
|
| 245 |
+
"rewards/margins": 0.2137639820575714,
|
| 246 |
+
"rewards/rejected": -0.025361843407154083,
|
| 247 |
+
"step": 350
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.94,
|
| 251 |
+
"learning_rate": 3.816155988857939e-07,
|
| 252 |
+
"logps/chosen": -68.25212097167969,
|
| 253 |
+
"logps/rejected": -69.1668930053711,
|
| 254 |
+
"loss": 0.6273,
|
| 255 |
+
"losses/dpo": 0.6162423491477966,
|
| 256 |
+
"losses/sft": 1.5002387762069702,
|
| 257 |
+
"losses/total": 0.6162423491477966,
|
| 258 |
+
"ref_logps/chosen": -70.23272705078125,
|
| 259 |
+
"ref_logps/rejected": -68.74607849121094,
|
| 260 |
+
"rewards/accuracies": 0.6295000314712524,
|
| 261 |
+
"rewards/chosen": 0.19805949926376343,
|
| 262 |
+
"rewards/margins": 0.24014097452163696,
|
| 263 |
+
"rewards/rejected": -0.04208146035671234,
|
| 264 |
+
"step": 375
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 1.0,
|
| 268 |
+
"learning_rate": 3.700092850510678e-07,
|
| 269 |
+
"logps/chosen": -70.95040893554688,
|
| 270 |
+
"logps/rejected": -72.96726989746094,
|
| 271 |
+
"loss": 0.6195,
|
| 272 |
+
"losses/dpo": 0.6564822196960449,
|
| 273 |
+
"losses/sft": 1.560630440711975,
|
| 274 |
+
"losses/total": 0.6564822196960449,
|
| 275 |
+
"ref_logps/chosen": -72.91325378417969,
|
| 276 |
+
"ref_logps/rejected": -72.14061737060547,
|
| 277 |
+
"rewards/accuracies": 0.6350000500679016,
|
| 278 |
+
"rewards/chosen": 0.19628457725048065,
|
| 279 |
+
"rewards/margins": 0.2789497375488281,
|
| 280 |
+
"rewards/rejected": -0.08266513794660568,
|
| 281 |
+
"step": 400
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 1.06,
|
| 285 |
+
"learning_rate": 3.5840297121634165e-07,
|
| 286 |
+
"logps/chosen": -67.00164794921875,
|
| 287 |
+
"logps/rejected": -69.44139099121094,
|
| 288 |
+
"loss": 0.6167,
|
| 289 |
+
"losses/dpo": 0.6204876899719238,
|
| 290 |
+
"losses/sft": 1.4992446899414062,
|
| 291 |
+
"losses/total": 0.6204876899719238,
|
| 292 |
+
"ref_logps/chosen": -68.97762298583984,
|
| 293 |
+
"ref_logps/rejected": -68.63683319091797,
|
| 294 |
+
"rewards/accuracies": 0.6190000772476196,
|
| 295 |
+
"rewards/chosen": 0.19759786128997803,
|
| 296 |
+
"rewards/margins": 0.27805399894714355,
|
| 297 |
+
"rewards/rejected": -0.08045615255832672,
|
| 298 |
+
"step": 425
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 1.13,
|
| 302 |
+
"learning_rate": 3.4679665738161556e-07,
|
| 303 |
+
"logps/chosen": -69.90442657470703,
|
| 304 |
+
"logps/rejected": -72.69268035888672,
|
| 305 |
+
"loss": 0.6054,
|
| 306 |
+
"losses/dpo": 0.598928689956665,
|
| 307 |
+
"losses/sft": 1.5615432262420654,
|
| 308 |
+
"losses/total": 0.598928689956665,
|
| 309 |
+
"ref_logps/chosen": -71.99002838134766,
|
| 310 |
+
"ref_logps/rejected": -71.58721923828125,
|
| 311 |
+
"rewards/accuracies": 0.6365000009536743,
|
| 312 |
+
"rewards/chosen": 0.20855939388275146,
|
| 313 |
+
"rewards/margins": 0.3191070556640625,
|
| 314 |
+
"rewards/rejected": -0.11054765433073044,
|
| 315 |
+
"step": 450
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 1.19,
|
| 319 |
+
"learning_rate": 3.351903435468895e-07,
|
| 320 |
+
"logps/chosen": -68.88660430908203,
|
| 321 |
+
"logps/rejected": -70.41629791259766,
|
| 322 |
+
"loss": 0.5879,
|
| 323 |
+
"losses/dpo": 0.5859370231628418,
|
| 324 |
+
"losses/sft": 1.5487860441207886,
|
| 325 |
+
"losses/total": 0.5859370231628418,
|
| 326 |
+
"ref_logps/chosen": -71.17853546142578,
|
| 327 |
+
"ref_logps/rejected": -69.02958679199219,
|
| 328 |
+
"rewards/accuracies": 0.6640000343322754,
|
| 329 |
+
"rewards/chosen": 0.22919251024723053,
|
| 330 |
+
"rewards/margins": 0.36786285042762756,
|
| 331 |
+
"rewards/rejected": -0.13867038488388062,
|
| 332 |
+
"step": 475
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 1.25,
|
| 336 |
+
"learning_rate": 3.2358402971216343e-07,
|
| 337 |
+
"logps/chosen": -69.55220031738281,
|
| 338 |
+
"logps/rejected": -70.18158721923828,
|
| 339 |
+
"loss": 0.6038,
|
| 340 |
+
"losses/dpo": 0.6201021075248718,
|
| 341 |
+
"losses/sft": 1.5225414037704468,
|
| 342 |
+
"losses/total": 0.6201021075248718,
|
| 343 |
+
"ref_logps/chosen": -71.74359130859375,
|
| 344 |
+
"ref_logps/rejected": -68.91885375976562,
|
| 345 |
+
"rewards/accuracies": 0.6455000042915344,
|
| 346 |
+
"rewards/chosen": 0.2191377878189087,
|
| 347 |
+
"rewards/margins": 0.34541237354278564,
|
| 348 |
+
"rewards/rejected": -0.12627457082271576,
|
| 349 |
+
"step": 500
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"epoch": 1.31,
|
| 353 |
+
"learning_rate": 3.1197771587743734e-07,
|
| 354 |
+
"logps/chosen": -69.61231231689453,
|
| 355 |
+
"logps/rejected": -72.09597778320312,
|
| 356 |
+
"loss": 0.5827,
|
| 357 |
+
"losses/dpo": 0.5971881747245789,
|
| 358 |
+
"losses/sft": 1.4868888854980469,
|
| 359 |
+
"losses/total": 0.5971881747245789,
|
| 360 |
+
"ref_logps/chosen": -71.90864562988281,
|
| 361 |
+
"ref_logps/rejected": -70.28910827636719,
|
| 362 |
+
"rewards/accuracies": 0.6584999561309814,
|
| 363 |
+
"rewards/chosen": 0.22963352501392365,
|
| 364 |
+
"rewards/margins": 0.41032031178474426,
|
| 365 |
+
"rewards/rejected": -0.18068677186965942,
|
| 366 |
+
"step": 525
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 1.38,
|
| 370 |
+
"learning_rate": 3.003714020427112e-07,
|
| 371 |
+
"logps/chosen": -69.58737182617188,
|
| 372 |
+
"logps/rejected": -70.84317016601562,
|
| 373 |
+
"loss": 0.5868,
|
| 374 |
+
"losses/dpo": 0.5683429837226868,
|
| 375 |
+
"losses/sft": 1.5186117887496948,
|
| 376 |
+
"losses/total": 0.5683429837226868,
|
| 377 |
+
"ref_logps/chosen": -71.8995590209961,
|
| 378 |
+
"ref_logps/rejected": -68.87027740478516,
|
| 379 |
+
"rewards/accuracies": 0.6439999938011169,
|
| 380 |
+
"rewards/chosen": 0.23121845722198486,
|
| 381 |
+
"rewards/margins": 0.42850860953330994,
|
| 382 |
+
"rewards/rejected": -0.19729015231132507,
|
| 383 |
+
"step": 550
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 1.44,
|
| 387 |
+
"learning_rate": 2.887650882079851e-07,
|
| 388 |
+
"logps/chosen": -69.75614166259766,
|
| 389 |
+
"logps/rejected": -71.54226684570312,
|
| 390 |
+
"loss": 0.5853,
|
| 391 |
+
"losses/dpo": 0.573384165763855,
|
| 392 |
+
"losses/sft": 1.506340742111206,
|
| 393 |
+
"losses/total": 0.573384165763855,
|
| 394 |
+
"ref_logps/chosen": -71.97299194335938,
|
| 395 |
+
"ref_logps/rejected": -69.5394287109375,
|
| 396 |
+
"rewards/accuracies": 0.6439999938011169,
|
| 397 |
+
"rewards/chosen": 0.2216847836971283,
|
| 398 |
+
"rewards/margins": 0.4219689667224884,
|
| 399 |
+
"rewards/rejected": -0.2002841979265213,
|
| 400 |
+
"step": 575
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 1.5,
|
| 404 |
+
"learning_rate": 2.77158774373259e-07,
|
| 405 |
+
"logps/chosen": -67.17892456054688,
|
| 406 |
+
"logps/rejected": -72.1897201538086,
|
| 407 |
+
"loss": 0.5811,
|
| 408 |
+
"losses/dpo": 0.5949756503105164,
|
| 409 |
+
"losses/sft": 1.5656284093856812,
|
| 410 |
+
"losses/total": 0.5949756503105164,
|
| 411 |
+
"ref_logps/chosen": -69.0622329711914,
|
| 412 |
+
"ref_logps/rejected": -69.59004974365234,
|
| 413 |
+
"rewards/accuracies": 0.658500075340271,
|
| 414 |
+
"rewards/chosen": 0.18833142518997192,
|
| 415 |
+
"rewards/margins": 0.4482985734939575,
|
| 416 |
+
"rewards/rejected": -0.2599670886993408,
|
| 417 |
+
"step": 600
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 1.56,
|
| 421 |
+
"learning_rate": 2.65552460538533e-07,
|
| 422 |
+
"logps/chosen": -67.06172180175781,
|
| 423 |
+
"logps/rejected": -71.94342041015625,
|
| 424 |
+
"loss": 0.5723,
|
| 425 |
+
"losses/dpo": 0.55665123462677,
|
| 426 |
+
"losses/sft": 1.5321460962295532,
|
| 427 |
+
"losses/total": 0.55665123462677,
|
| 428 |
+
"ref_logps/chosen": -68.80352783203125,
|
| 429 |
+
"ref_logps/rejected": -69.0389175415039,
|
| 430 |
+
"rewards/accuracies": 0.6610000133514404,
|
| 431 |
+
"rewards/chosen": 0.17417989671230316,
|
| 432 |
+
"rewards/margins": 0.46463003754615784,
|
| 433 |
+
"rewards/rejected": -0.29045018553733826,
|
| 434 |
+
"step": 625
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 1.63,
|
| 438 |
+
"learning_rate": 2.539461467038069e-07,
|
| 439 |
+
"logps/chosen": -71.48672485351562,
|
| 440 |
+
"logps/rejected": -74.49577331542969,
|
| 441 |
+
"loss": 0.5551,
|
| 442 |
+
"losses/dpo": 0.5403231382369995,
|
| 443 |
+
"losses/sft": 1.5089354515075684,
|
| 444 |
+
"losses/total": 0.5403231382369995,
|
| 445 |
+
"ref_logps/chosen": -73.56059265136719,
|
| 446 |
+
"ref_logps/rejected": -71.07339477539062,
|
| 447 |
+
"rewards/accuracies": 0.6800000071525574,
|
| 448 |
+
"rewards/chosen": 0.20738649368286133,
|
| 449 |
+
"rewards/margins": 0.5496238470077515,
|
| 450 |
+
"rewards/rejected": -0.34223735332489014,
|
| 451 |
+
"step": 650
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 1.69,
|
| 455 |
+
"learning_rate": 2.4233983286908075e-07,
|
| 456 |
+
"logps/chosen": -70.11666107177734,
|
| 457 |
+
"logps/rejected": -72.92801666259766,
|
| 458 |
+
"loss": 0.5697,
|
| 459 |
+
"losses/dpo": 0.5605096817016602,
|
| 460 |
+
"losses/sft": 1.5102070569992065,
|
| 461 |
+
"losses/total": 0.5605096817016602,
|
| 462 |
+
"ref_logps/chosen": -71.93968200683594,
|
| 463 |
+
"ref_logps/rejected": -69.70313262939453,
|
| 464 |
+
"rewards/accuracies": 0.6594999432563782,
|
| 465 |
+
"rewards/chosen": 0.18230296671390533,
|
| 466 |
+
"rewards/margins": 0.5047909021377563,
|
| 467 |
+
"rewards/rejected": -0.32248786091804504,
|
| 468 |
+
"step": 675
|
| 469 |
+
},
|
| 470 |
+
{
|
| 471 |
+
"epoch": 1.75,
|
| 472 |
+
"learning_rate": 2.3073351903435469e-07,
|
| 473 |
+
"logps/chosen": -70.7259750366211,
|
| 474 |
+
"logps/rejected": -74.96146392822266,
|
| 475 |
+
"loss": 0.5596,
|
| 476 |
+
"losses/dpo": 0.5713181495666504,
|
| 477 |
+
"losses/sft": 1.501438021659851,
|
| 478 |
+
"losses/total": 0.5713181495666504,
|
| 479 |
+
"ref_logps/chosen": -72.50656127929688,
|
| 480 |
+
"ref_logps/rejected": -71.40023803710938,
|
| 481 |
+
"rewards/accuracies": 0.6769999861717224,
|
| 482 |
+
"rewards/chosen": 0.1780581921339035,
|
| 483 |
+
"rewards/margins": 0.5341811180114746,
|
| 484 |
+
"rewards/rejected": -0.3561229705810547,
|
| 485 |
+
"step": 700
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 1.81,
|
| 489 |
+
"learning_rate": 2.191272051996286e-07,
|
| 490 |
+
"logps/chosen": -71.7147445678711,
|
| 491 |
+
"logps/rejected": -74.73912048339844,
|
| 492 |
+
"loss": 0.5536,
|
| 493 |
+
"losses/dpo": 0.5598438382148743,
|
| 494 |
+
"losses/sft": 1.6115312576293945,
|
| 495 |
+
"losses/total": 0.5598438382148743,
|
| 496 |
+
"ref_logps/chosen": -73.46680450439453,
|
| 497 |
+
"ref_logps/rejected": -70.61848449707031,
|
| 498 |
+
"rewards/accuracies": 0.6815000772476196,
|
| 499 |
+
"rewards/chosen": 0.17520827054977417,
|
| 500 |
+
"rewards/margins": 0.5872728228569031,
|
| 501 |
+
"rewards/rejected": -0.4120645225048065,
|
| 502 |
+
"step": 725
|
| 503 |
+
},
|
| 504 |
+
{
|
| 505 |
+
"epoch": 1.88,
|
| 506 |
+
"learning_rate": 2.075208913649025e-07,
|
| 507 |
+
"logps/chosen": -70.7854232788086,
|
| 508 |
+
"logps/rejected": -74.75676727294922,
|
| 509 |
+
"loss": 0.55,
|
| 510 |
+
"losses/dpo": 0.5355216860771179,
|
| 511 |
+
"losses/sft": 1.5508781671524048,
|
| 512 |
+
"losses/total": 0.5355216860771179,
|
| 513 |
+
"ref_logps/chosen": -72.55353546142578,
|
| 514 |
+
"ref_logps/rejected": -70.52770233154297,
|
| 515 |
+
"rewards/accuracies": 0.6890000104904175,
|
| 516 |
+
"rewards/chosen": 0.1768111288547516,
|
| 517 |
+
"rewards/margins": 0.5997176170349121,
|
| 518 |
+
"rewards/rejected": -0.4229064881801605,
|
| 519 |
+
"step": 750
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 1.94,
|
| 523 |
+
"learning_rate": 1.959145775301764e-07,
|
| 524 |
+
"logps/chosen": -69.92965698242188,
|
| 525 |
+
"logps/rejected": -75.43944549560547,
|
| 526 |
+
"loss": 0.5567,
|
| 527 |
+
"losses/dpo": 0.5669773817062378,
|
| 528 |
+
"losses/sft": 1.5475414991378784,
|
| 529 |
+
"losses/total": 0.5669773817062378,
|
| 530 |
+
"ref_logps/chosen": -71.25016021728516,
|
| 531 |
+
"ref_logps/rejected": -70.77532958984375,
|
| 532 |
+
"rewards/accuracies": 0.6794999837875366,
|
| 533 |
+
"rewards/chosen": 0.13204967975616455,
|
| 534 |
+
"rewards/margins": 0.5984623432159424,
|
| 535 |
+
"rewards/rejected": -0.4664126932621002,
|
| 536 |
+
"step": 775
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 2.0,
|
| 540 |
+
"learning_rate": 1.8430826369545033e-07,
|
| 541 |
+
"logps/chosen": -70.1933364868164,
|
| 542 |
+
"logps/rejected": -74.6698226928711,
|
| 543 |
+
"loss": 0.5486,
|
| 544 |
+
"losses/dpo": 0.543175458908081,
|
| 545 |
+
"losses/sft": 1.5320526361465454,
|
| 546 |
+
"losses/total": 0.543175458908081,
|
| 547 |
+
"ref_logps/chosen": -71.89373016357422,
|
| 548 |
+
"ref_logps/rejected": -70.11508178710938,
|
| 549 |
+
"rewards/accuracies": 0.6875,
|
| 550 |
+
"rewards/chosen": 0.17003829777240753,
|
| 551 |
+
"rewards/margins": 0.6255122423171997,
|
| 552 |
+
"rewards/rejected": -0.455473929643631,
|
| 553 |
+
"step": 800
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 2.07,
|
| 557 |
+
"learning_rate": 1.7270194986072424e-07,
|
| 558 |
+
"logps/chosen": -68.841064453125,
|
| 559 |
+
"logps/rejected": -75.60282897949219,
|
| 560 |
+
"loss": 0.5496,
|
| 561 |
+
"losses/dpo": 0.5592978596687317,
|
| 562 |
+
"losses/sft": 1.591374158859253,
|
| 563 |
+
"losses/total": 0.5592978596687317,
|
| 564 |
+
"ref_logps/chosen": -70.10298156738281,
|
| 565 |
+
"ref_logps/rejected": -70.73596954345703,
|
| 566 |
+
"rewards/accuracies": 0.684499979019165,
|
| 567 |
+
"rewards/chosen": 0.12619122862815857,
|
| 568 |
+
"rewards/margins": 0.6128779053688049,
|
| 569 |
+
"rewards/rejected": -0.48668670654296875,
|
| 570 |
+
"step": 825
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 2.13,
|
| 574 |
+
"learning_rate": 1.6109563602599812e-07,
|
| 575 |
+
"logps/chosen": -70.97602081298828,
|
| 576 |
+
"logps/rejected": -75.59082794189453,
|
| 577 |
+
"loss": 0.5226,
|
| 578 |
+
"losses/dpo": 0.5137518048286438,
|
| 579 |
+
"losses/sft": 1.4946039915084839,
|
| 580 |
+
"losses/total": 0.5137518048286438,
|
| 581 |
+
"ref_logps/chosen": -72.67564392089844,
|
| 582 |
+
"ref_logps/rejected": -70.24249267578125,
|
| 583 |
+
"rewards/accuracies": 0.7165000438690186,
|
| 584 |
+
"rewards/chosen": 0.169962078332901,
|
| 585 |
+
"rewards/margins": 0.704794704914093,
|
| 586 |
+
"rewards/rejected": -0.5348325967788696,
|
| 587 |
+
"step": 850
|
| 588 |
+
},
|
| 589 |
+
{
|
| 590 |
+
"epoch": 2.19,
|
| 591 |
+
"learning_rate": 1.4948932219127206e-07,
|
| 592 |
+
"logps/chosen": -69.44739532470703,
|
| 593 |
+
"logps/rejected": -74.73712158203125,
|
| 594 |
+
"loss": 0.5483,
|
| 595 |
+
"losses/dpo": 0.554201602935791,
|
| 596 |
+
"losses/sft": 1.5423518419265747,
|
| 597 |
+
"losses/total": 0.554201602935791,
|
| 598 |
+
"ref_logps/chosen": -70.63737487792969,
|
| 599 |
+
"ref_logps/rejected": -69.3394546508789,
|
| 600 |
+
"rewards/accuracies": 0.6875,
|
| 601 |
+
"rewards/chosen": 0.11899794638156891,
|
| 602 |
+
"rewards/margins": 0.6587647795677185,
|
| 603 |
+
"rewards/rejected": -0.5397669076919556,
|
| 604 |
+
"step": 875
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 2.25,
|
| 608 |
+
"learning_rate": 1.3788300835654597e-07,
|
| 609 |
+
"logps/chosen": -67.69676971435547,
|
| 610 |
+
"logps/rejected": -72.41621398925781,
|
| 611 |
+
"loss": 0.554,
|
| 612 |
+
"losses/dpo": 0.536446750164032,
|
| 613 |
+
"losses/sft": 1.52887761592865,
|
| 614 |
+
"losses/total": 0.536446750164032,
|
| 615 |
+
"ref_logps/chosen": -68.82073974609375,
|
| 616 |
+
"ref_logps/rejected": -67.15235137939453,
|
| 617 |
+
"rewards/accuracies": 0.6720000505447388,
|
| 618 |
+
"rewards/chosen": 0.11239679157733917,
|
| 619 |
+
"rewards/margins": 0.6387830972671509,
|
| 620 |
+
"rewards/rejected": -0.5263863205909729,
|
| 621 |
+
"step": 900
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 2.32,
|
| 625 |
+
"learning_rate": 1.2627669452181985e-07,
|
| 626 |
+
"logps/chosen": -68.61820983886719,
|
| 627 |
+
"logps/rejected": -75.03334045410156,
|
| 628 |
+
"loss": 0.547,
|
| 629 |
+
"losses/dpo": 0.5469151139259338,
|
| 630 |
+
"losses/sft": 1.524001121520996,
|
| 631 |
+
"losses/total": 0.5469151139259338,
|
| 632 |
+
"ref_logps/chosen": -69.68536376953125,
|
| 633 |
+
"ref_logps/rejected": -69.33779907226562,
|
| 634 |
+
"rewards/accuracies": 0.6725000143051147,
|
| 635 |
+
"rewards/chosen": 0.10671478509902954,
|
| 636 |
+
"rewards/margins": 0.6762691736221313,
|
| 637 |
+
"rewards/rejected": -0.5695543885231018,
|
| 638 |
+
"step": 925
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 2.38,
|
| 642 |
+
"learning_rate": 1.1467038068709377e-07,
|
| 643 |
+
"logps/chosen": -71.48025512695312,
|
| 644 |
+
"logps/rejected": -76.9962387084961,
|
| 645 |
+
"loss": 0.5328,
|
| 646 |
+
"losses/dpo": 0.5382718443870544,
|
| 647 |
+
"losses/sft": 1.5356690883636475,
|
| 648 |
+
"losses/total": 0.5382718443870544,
|
| 649 |
+
"ref_logps/chosen": -72.66265869140625,
|
| 650 |
+
"ref_logps/rejected": -70.87562561035156,
|
| 651 |
+
"rewards/accuracies": 0.7055000066757202,
|
| 652 |
+
"rewards/chosen": 0.11823976784944534,
|
| 653 |
+
"rewards/margins": 0.7303012013435364,
|
| 654 |
+
"rewards/rejected": -0.6120614409446716,
|
| 655 |
+
"step": 950
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 2.44,
|
| 659 |
+
"learning_rate": 1.0306406685236768e-07,
|
| 660 |
+
"logps/chosen": -68.71895599365234,
|
| 661 |
+
"logps/rejected": -74.29911804199219,
|
| 662 |
+
"loss": 0.5524,
|
| 663 |
+
"losses/dpo": 0.5623547434806824,
|
| 664 |
+
"losses/sft": 1.6026860475540161,
|
| 665 |
+
"losses/total": 0.5623547434806824,
|
| 666 |
+
"ref_logps/chosen": -69.32998657226562,
|
| 667 |
+
"ref_logps/rejected": -68.43716430664062,
|
| 668 |
+
"rewards/accuracies": 0.6655000448226929,
|
| 669 |
+
"rewards/chosen": 0.06110435351729393,
|
| 670 |
+
"rewards/margins": 0.6473007202148438,
|
| 671 |
+
"rewards/rejected": -0.5861963629722595,
|
| 672 |
+
"step": 975
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 2.5,
|
| 676 |
+
"learning_rate": 9.14577530176416e-08,
|
| 677 |
+
"logps/chosen": -69.40322875976562,
|
| 678 |
+
"logps/rejected": -73.88810729980469,
|
| 679 |
+
"loss": 0.5513,
|
| 680 |
+
"losses/dpo": 0.5662988424301147,
|
| 681 |
+
"losses/sft": 1.6082065105438232,
|
| 682 |
+
"losses/total": 0.5662988424301147,
|
| 683 |
+
"ref_logps/chosen": -70.1698226928711,
|
| 684 |
+
"ref_logps/rejected": -68.06956481933594,
|
| 685 |
+
"rewards/accuracies": 0.6759999990463257,
|
| 686 |
+
"rewards/chosen": 0.07665982842445374,
|
| 687 |
+
"rewards/margins": 0.6585137248039246,
|
| 688 |
+
"rewards/rejected": -0.5818539261817932,
|
| 689 |
+
"step": 1000
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 2.57,
|
| 693 |
+
"learning_rate": 7.98514391829155e-08,
|
| 694 |
+
"logps/chosen": -68.32543182373047,
|
| 695 |
+
"logps/rejected": -74.76167297363281,
|
| 696 |
+
"loss": 0.5433,
|
| 697 |
+
"losses/dpo": 0.5388572216033936,
|
| 698 |
+
"losses/sft": 1.5300703048706055,
|
| 699 |
+
"losses/total": 0.5388572216033936,
|
| 700 |
+
"ref_logps/chosen": -69.11152648925781,
|
| 701 |
+
"ref_logps/rejected": -68.69097900390625,
|
| 702 |
+
"rewards/accuracies": 0.6825000047683716,
|
| 703 |
+
"rewards/chosen": 0.07860930263996124,
|
| 704 |
+
"rewards/margins": 0.685679018497467,
|
| 705 |
+
"rewards/rejected": -0.6070696115493774,
|
| 706 |
+
"step": 1025
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 2.63,
|
| 710 |
+
"learning_rate": 6.824512534818941e-08,
|
| 711 |
+
"logps/chosen": -70.73451232910156,
|
| 712 |
+
"logps/rejected": -77.00275421142578,
|
| 713 |
+
"loss": 0.5239,
|
| 714 |
+
"losses/dpo": 0.5082178115844727,
|
| 715 |
+
"losses/sft": 1.4840093851089478,
|
| 716 |
+
"losses/total": 0.5082178115844727,
|
| 717 |
+
"ref_logps/chosen": -72.08134460449219,
|
| 718 |
+
"ref_logps/rejected": -70.49996948242188,
|
| 719 |
+
"rewards/accuracies": 0.7019999623298645,
|
| 720 |
+
"rewards/chosen": 0.13468389213085175,
|
| 721 |
+
"rewards/margins": 0.7849621772766113,
|
| 722 |
+
"rewards/rejected": -0.6502782702445984,
|
| 723 |
+
"step": 1050
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 2.69,
|
| 727 |
+
"learning_rate": 5.6638811513463324e-08,
|
| 728 |
+
"logps/chosen": -69.95764923095703,
|
| 729 |
+
"logps/rejected": -77.90116882324219,
|
| 730 |
+
"loss": 0.5327,
|
| 731 |
+
"losses/dpo": 0.5328630805015564,
|
| 732 |
+
"losses/sft": 1.6418886184692383,
|
| 733 |
+
"losses/total": 0.5328630805015564,
|
| 734 |
+
"ref_logps/chosen": -70.50801086425781,
|
| 735 |
+
"ref_logps/rejected": -71.11058807373047,
|
| 736 |
+
"rewards/accuracies": 0.6959999799728394,
|
| 737 |
+
"rewards/chosen": 0.05503645911812782,
|
| 738 |
+
"rewards/margins": 0.7340949773788452,
|
| 739 |
+
"rewards/rejected": -0.6790586113929749,
|
| 740 |
+
"step": 1075
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 2.75,
|
| 744 |
+
"learning_rate": 4.503249767873723e-08,
|
| 745 |
+
"logps/chosen": -69.83995819091797,
|
| 746 |
+
"logps/rejected": -75.7170639038086,
|
| 747 |
+
"loss": 0.5415,
|
| 748 |
+
"losses/dpo": 0.5642114281654358,
|
| 749 |
+
"losses/sft": 1.5595824718475342,
|
| 750 |
+
"losses/total": 0.5642114281654358,
|
| 751 |
+
"ref_logps/chosen": -70.96809387207031,
|
| 752 |
+
"ref_logps/rejected": -69.68138885498047,
|
| 753 |
+
"rewards/accuracies": 0.6990000009536743,
|
| 754 |
+
"rewards/chosen": 0.11281368136405945,
|
| 755 |
+
"rewards/margins": 0.7163800001144409,
|
| 756 |
+
"rewards/rejected": -0.6035662889480591,
|
| 757 |
+
"step": 1100
|
| 758 |
+
},
|
| 759 |
+
{
|
| 760 |
+
"epoch": 2.82,
|
| 761 |
+
"learning_rate": 3.3426183844011144e-08,
|
| 762 |
+
"logps/chosen": -72.0064697265625,
|
| 763 |
+
"logps/rejected": -75.64459228515625,
|
| 764 |
+
"loss": 0.553,
|
| 765 |
+
"losses/dpo": 0.6136656403541565,
|
| 766 |
+
"losses/sft": 1.6066731214523315,
|
| 767 |
+
"losses/total": 0.6136656403541565,
|
| 768 |
+
"ref_logps/chosen": -72.82428741455078,
|
| 769 |
+
"ref_logps/rejected": -69.51007843017578,
|
| 770 |
+
"rewards/accuracies": 0.6880000233650208,
|
| 771 |
+
"rewards/chosen": 0.08178197592496872,
|
| 772 |
+
"rewards/margins": 0.6952335834503174,
|
| 773 |
+
"rewards/rejected": -0.6134517192840576,
|
| 774 |
+
"step": 1125
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 2.88,
|
| 778 |
+
"learning_rate": 2.181987000928505e-08,
|
| 779 |
+
"logps/chosen": -71.19115447998047,
|
| 780 |
+
"logps/rejected": -74.97571563720703,
|
| 781 |
+
"loss": 0.5495,
|
| 782 |
+
"losses/dpo": 0.5555659532546997,
|
| 783 |
+
"losses/sft": 1.547566533088684,
|
| 784 |
+
"losses/total": 0.5555659532546997,
|
| 785 |
+
"ref_logps/chosen": -72.04399108886719,
|
| 786 |
+
"ref_logps/rejected": -68.92461395263672,
|
| 787 |
+
"rewards/accuracies": 0.6770000457763672,
|
| 788 |
+
"rewards/chosen": 0.08528263866901398,
|
| 789 |
+
"rewards/margins": 0.6903927326202393,
|
| 790 |
+
"rewards/rejected": -0.6051101088523865,
|
| 791 |
+
"step": 1150
|
| 792 |
+
},
|
| 793 |
+
{
|
| 794 |
+
"epoch": 2.94,
|
| 795 |
+
"learning_rate": 1.0213556174558959e-08,
|
| 796 |
+
"logps/chosen": -68.98538970947266,
|
| 797 |
+
"logps/rejected": -74.58392333984375,
|
| 798 |
+
"loss": 0.5254,
|
| 799 |
+
"losses/dpo": 0.49954432249069214,
|
| 800 |
+
"losses/sft": 1.4814612865447998,
|
| 801 |
+
"losses/total": 0.49954432249069214,
|
| 802 |
+
"ref_logps/chosen": -70.0972671508789,
|
| 803 |
+
"ref_logps/rejected": -68.21308898925781,
|
| 804 |
+
"rewards/accuracies": 0.6914999485015869,
|
| 805 |
+
"rewards/chosen": 0.11118759214878082,
|
| 806 |
+
"rewards/margins": 0.7482713460922241,
|
| 807 |
+
"rewards/rejected": -0.6370838284492493,
|
| 808 |
+
"step": 1175
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 3.0,
|
| 812 |
+
"step": 1197,
|
| 813 |
+
"total_flos": 0.0,
|
| 814 |
+
"train_loss": 0.5977537606095112,
|
| 815 |
+
"train_runtime": 15731.8751,
|
| 816 |
+
"train_samples_per_second": 6.094,
|
| 817 |
+
"train_steps_per_second": 0.076
|
| 818 |
+
}
|
| 819 |
+
],
|
| 820 |
+
"logging_steps": 25,
|
| 821 |
+
"max_steps": 1197,
|
| 822 |
+
"num_input_tokens_seen": 0,
|
| 823 |
+
"num_train_epochs": 3,
|
| 824 |
+
"save_steps": 500,
|
| 825 |
+
"stateful_callbacks": {},
|
| 826 |
+
"total_flos": 0.0,
|
| 827 |
+
"train_batch_size": 4,
|
| 828 |
+
"trial_name": null,
|
| 829 |
+
"trial_params": null
|
| 830 |
+
}
|