Commit
•
6123763
1
Parent(s):
27d0683
Upload folder using huggingface_hub
Browse files- README.md +9 -0
- adapter_config.json +26 -0
- adapter_model.bin +3 -0
- config.json +45 -0
- non_lora_trainables.bin +3 -0
- trainer_state.json +655 -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 0.4.0
|
adapter_config.json
ADDED
|
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"auto_mapping": null,
|
| 3 |
+
"base_model_name_or_path": "/home/ubuntu/LLaVA/checkpoints/llava-v1.5-13b-steer-lora-helpsteerformat-interleaf-lrv3067llava3067-likert-1222_interleavedgpt4v_responseonly-pure-merged",
|
| 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 |
+
"q_proj",
|
| 18 |
+
"o_proj",
|
| 19 |
+
"up_proj",
|
| 20 |
+
"down_proj",
|
| 21 |
+
"gate_proj",
|
| 22 |
+
"k_proj",
|
| 23 |
+
"v_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 sha256:31d06a25c8e47fc68831a99c3d505fa3d8cdbaab1427da085cd253c047b76f1f
|
| 3 |
+
size 1001584333
|
config.json
ADDED
|
@@ -0,0 +1,45 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "/home/ubuntu/LLaVA/checkpoints/llava-v1.5-13b-steer-lora-helpsteerformat-interleaf-lrv3067llava3067-likert-1222_interleavedgpt4v_responseonly-pure-merged",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"LlavaLlamaForCausalLM"
|
| 5 |
+
],
|
| 6 |
+
"bos_token_id": 1,
|
| 7 |
+
"eos_token_id": 2,
|
| 8 |
+
"freeze_mm_mlp_adapter": false,
|
| 9 |
+
"freeze_mm_vision_resampler": false,
|
| 10 |
+
"hidden_act": "silu",
|
| 11 |
+
"hidden_size": 5120,
|
| 12 |
+
"image_aspect_ratio": "pad",
|
| 13 |
+
"initializer_range": 0.02,
|
| 14 |
+
"intermediate_size": 13824,
|
| 15 |
+
"max_length": 4096,
|
| 16 |
+
"max_position_embeddings": 4096,
|
| 17 |
+
"mm_hidden_size": 1024,
|
| 18 |
+
"mm_projector_lr": 2e-05,
|
| 19 |
+
"mm_projector_type": "mlp2x_gelu",
|
| 20 |
+
"mm_resampler_type": null,
|
| 21 |
+
"mm_use_im_patch_token": false,
|
| 22 |
+
"mm_use_im_start_end": false,
|
| 23 |
+
"mm_vision_select_feature": "patch",
|
| 24 |
+
"mm_vision_select_layer": -2,
|
| 25 |
+
"mm_vision_tower": "openai/clip-vit-large-patch14-336",
|
| 26 |
+
"model_type": "llava",
|
| 27 |
+
"num_attention_heads": 40,
|
| 28 |
+
"num_hidden_layers": 40,
|
| 29 |
+
"num_key_value_heads": 40,
|
| 30 |
+
"pad_token_id": 0,
|
| 31 |
+
"pretraining_tp": 1,
|
| 32 |
+
"rms_norm_eps": 1e-05,
|
| 33 |
+
"rope_scaling": null,
|
| 34 |
+
"tie_word_embeddings": false,
|
| 35 |
+
"tokenizer_model_max_length": 2048,
|
| 36 |
+
"tokenizer_padding_side": "right",
|
| 37 |
+
"torch_dtype": "float16",
|
| 38 |
+
"transformers_version": "4.31.0",
|
| 39 |
+
"tune_mm_mlp_adapter": false,
|
| 40 |
+
"tune_mm_vision_resampler": false,
|
| 41 |
+
"unfreeze_mm_vision_tower": false,
|
| 42 |
+
"use_cache": true,
|
| 43 |
+
"use_mm_proj": true,
|
| 44 |
+
"vocab_size": 32000
|
| 45 |
+
}
|
non_lora_trainables.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d08a173a526e8a979404fc833be6312a1638363d2545270d93170a3e3ebf41c8
|
| 3 |
+
size 62936807
|
trainer_state.json
ADDED
|
@@ -0,0 +1,655 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.8,
|
| 5 |
+
"global_step": 21,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0,
|
| 12 |
+
"logps_train/chosen": -100.49485778808594,
|
| 13 |
+
"logps_train/ref_chosen": -100.5,
|
| 14 |
+
"logps_train/ref_rejected": -105.0,
|
| 15 |
+
"logps_train/rejected": -104.80752563476562,
|
| 16 |
+
"rewards_train/accuracies": 0.46875,
|
| 17 |
+
"rewards_train/chosen": -0.020498279482126236,
|
| 18 |
+
"rewards_train/margins": -0.021032828837633133,
|
| 19 |
+
"rewards_train/rejected": 0.000534549355506897,
|
| 20 |
+
"step": 0
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0,
|
| 24 |
+
"logps_train/chosen": -89.90950012207031,
|
| 25 |
+
"logps_train/ref_chosen": -90.0,
|
| 26 |
+
"logps_train/ref_rejected": -101.0,
|
| 27 |
+
"logps_train/rejected": -100.86872863769531,
|
| 28 |
+
"rewards_train/accuracies": 0.5,
|
| 29 |
+
"rewards_train/chosen": 0.0191262885928154,
|
| 30 |
+
"rewards_train/margins": 0.0020734891295433044,
|
| 31 |
+
"rewards_train/rejected": 0.017052799463272095,
|
| 32 |
+
"step": 0
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.13,
|
| 36 |
+
"learning_rate": 5e-06,
|
| 37 |
+
"loss": 0.7043,
|
| 38 |
+
"step": 1
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.13,
|
| 42 |
+
"logps_train/chosen": -104.38166809082031,
|
| 43 |
+
"logps_train/ref_chosen": -104.5,
|
| 44 |
+
"logps_train/ref_rejected": -98.0,
|
| 45 |
+
"logps_train/rejected": -97.89907836914062,
|
| 46 |
+
"rewards_train/accuracies": 0.515625,
|
| 47 |
+
"rewards_train/chosen": 0.03059956058859825,
|
| 48 |
+
"rewards_train/margins": 0.004916800186038017,
|
| 49 |
+
"rewards_train/rejected": 0.025682760402560234,
|
| 50 |
+
"step": 1
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.13,
|
| 54 |
+
"logps_train/chosen": -94.88069152832031,
|
| 55 |
+
"logps_train/ref_chosen": -95.0,
|
| 56 |
+
"logps_train/ref_rejected": -95.0,
|
| 57 |
+
"logps_train/rejected": -95.29293823242188,
|
| 58 |
+
"rewards_train/accuracies": 0.546875,
|
| 59 |
+
"rewards_train/chosen": -0.009438544511795044,
|
| 60 |
+
"rewards_train/margins": -0.020688317716121674,
|
| 61 |
+
"rewards_train/rejected": 0.01124977320432663,
|
| 62 |
+
"step": 1
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.27,
|
| 66 |
+
"learning_rate": 4.9692208514878445e-06,
|
| 67 |
+
"loss": 0.7061,
|
| 68 |
+
"step": 2
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.27,
|
| 72 |
+
"logps_train/chosen": -82.53303527832031,
|
| 73 |
+
"logps_train/ref_chosen": -82.5,
|
| 74 |
+
"logps_train/ref_rejected": -89.0,
|
| 75 |
+
"logps_train/rejected": -88.99131774902344,
|
| 76 |
+
"rewards_train/accuracies": 0.484375,
|
| 77 |
+
"rewards_train/chosen": 0.006856412626802921,
|
| 78 |
+
"rewards_train/margins": -0.001938466913998127,
|
| 79 |
+
"rewards_train/rejected": 0.008794879540801048,
|
| 80 |
+
"step": 2
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.27,
|
| 84 |
+
"logps_train/chosen": -100.22808837890625,
|
| 85 |
+
"logps_train/ref_chosen": -100.0,
|
| 86 |
+
"logps_train/ref_rejected": -106.0,
|
| 87 |
+
"logps_train/rejected": -106.06494140625,
|
| 88 |
+
"rewards_train/accuracies": 0.46875,
|
| 89 |
+
"rewards_train/chosen": -0.033964600414037704,
|
| 90 |
+
"rewards_train/margins": -0.008083481341600418,
|
| 91 |
+
"rewards_train/rejected": -0.025881119072437286,
|
| 92 |
+
"step": 2
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.4,
|
| 96 |
+
"learning_rate": 4.8776412907378845e-06,
|
| 97 |
+
"loss": 0.702,
|
| 98 |
+
"step": 3
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.4,
|
| 102 |
+
"logps_train/chosen": -85.86089324951172,
|
| 103 |
+
"logps_train/ref_chosen": -86.0,
|
| 104 |
+
"logps_train/ref_rejected": -90.0,
|
| 105 |
+
"logps_train/rejected": -89.91110229492188,
|
| 106 |
+
"rewards_train/accuracies": 0.3125,
|
| 107 |
+
"rewards_train/chosen": -0.007595188915729523,
|
| 108 |
+
"rewards_train/margins": -0.07023129612207413,
|
| 109 |
+
"rewards_train/rejected": 0.0626361072063446,
|
| 110 |
+
"step": 3
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.4,
|
| 114 |
+
"logps_train/chosen": -92.3138427734375,
|
| 115 |
+
"logps_train/ref_chosen": -92.0,
|
| 116 |
+
"logps_train/ref_rejected": -77.5,
|
| 117 |
+
"logps_train/rejected": -77.53357696533203,
|
| 118 |
+
"rewards_train/accuracies": 0.453125,
|
| 119 |
+
"rewards_train/chosen": -0.0021358560770750046,
|
| 120 |
+
"rewards_train/margins": -0.025264138355851173,
|
| 121 |
+
"rewards_train/rejected": 0.02312828227877617,
|
| 122 |
+
"step": 3
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.53,
|
| 126 |
+
"learning_rate": 4.72751631047092e-06,
|
| 127 |
+
"loss": 0.7233,
|
| 128 |
+
"step": 4
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.53,
|
| 132 |
+
"logps_train/chosen": -88.56388854980469,
|
| 133 |
+
"logps_train/ref_chosen": -88.5,
|
| 134 |
+
"logps_train/ref_rejected": -91.5,
|
| 135 |
+
"logps_train/rejected": -91.46757507324219,
|
| 136 |
+
"rewards_train/accuracies": 0.40625,
|
| 137 |
+
"rewards_train/chosen": 0.0178141500800848,
|
| 138 |
+
"rewards_train/margins": -0.01190277375280857,
|
| 139 |
+
"rewards_train/rejected": 0.02971692383289337,
|
| 140 |
+
"step": 4
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.53,
|
| 144 |
+
"logps_train/chosen": -104.78536224365234,
|
| 145 |
+
"logps_train/ref_chosen": -105.0,
|
| 146 |
+
"logps_train/ref_rejected": -104.5,
|
| 147 |
+
"logps_train/rejected": -104.33482360839844,
|
| 148 |
+
"rewards_train/accuracies": 0.59375,
|
| 149 |
+
"rewards_train/chosen": 0.07069776952266693,
|
| 150 |
+
"rewards_train/margins": 0.03803768381476402,
|
| 151 |
+
"rewards_train/rejected": 0.03266008570790291,
|
| 152 |
+
"step": 4
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.67,
|
| 156 |
+
"learning_rate": 4.522542485937369e-06,
|
| 157 |
+
"loss": 0.6957,
|
| 158 |
+
"step": 5
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.67,
|
| 162 |
+
"logps_train/chosen": -109.24040985107422,
|
| 163 |
+
"logps_train/ref_chosen": -109.0,
|
| 164 |
+
"logps_train/ref_rejected": -99.5,
|
| 165 |
+
"logps_train/rejected": -99.40049743652344,
|
| 166 |
+
"rewards_train/accuracies": 0.59375,
|
| 167 |
+
"rewards_train/chosen": 0.030185997486114502,
|
| 168 |
+
"rewards_train/margins": 0.04000556468963623,
|
| 169 |
+
"rewards_train/rejected": -0.009819567203521729,
|
| 170 |
+
"step": 5
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.67,
|
| 174 |
+
"logps_train/chosen": -82.44449615478516,
|
| 175 |
+
"logps_train/ref_chosen": -82.5,
|
| 176 |
+
"logps_train/ref_rejected": -80.0,
|
| 177 |
+
"logps_train/rejected": -79.84416198730469,
|
| 178 |
+
"rewards_train/accuracies": 0.5625,
|
| 179 |
+
"rewards_train/chosen": 0.07816548645496368,
|
| 180 |
+
"rewards_train/margins": 0.038515977561473846,
|
| 181 |
+
"rewards_train/rejected": 0.03964950889348984,
|
| 182 |
+
"step": 5
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.8,
|
| 186 |
+
"learning_rate": 4.267766952966369e-06,
|
| 187 |
+
"loss": 0.6776,
|
| 188 |
+
"step": 6
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.8,
|
| 192 |
+
"logps_train/chosen": -92.00798034667969,
|
| 193 |
+
"logps_train/ref_chosen": -92.0,
|
| 194 |
+
"logps_train/ref_rejected": -89.5,
|
| 195 |
+
"logps_train/rejected": -89.4742660522461,
|
| 196 |
+
"rewards_train/accuracies": 0.59375,
|
| 197 |
+
"rewards_train/chosen": 0.10367429256439209,
|
| 198 |
+
"rewards_train/margins": 0.02098412811756134,
|
| 199 |
+
"rewards_train/rejected": 0.08269016444683075,
|
| 200 |
+
"step": 6
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.8,
|
| 204 |
+
"logps_train/chosen": -88.19847106933594,
|
| 205 |
+
"logps_train/ref_chosen": -88.5,
|
| 206 |
+
"logps_train/ref_rejected": -94.0,
|
| 207 |
+
"logps_train/rejected": -93.69485473632812,
|
| 208 |
+
"rewards_train/accuracies": 0.53125,
|
| 209 |
+
"rewards_train/chosen": 0.054087504744529724,
|
| 210 |
+
"rewards_train/margins": 0.018747031688690186,
|
| 211 |
+
"rewards_train/rejected": 0.03534047305583954,
|
| 212 |
+
"step": 6
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.93,
|
| 216 |
+
"learning_rate": 3.969463130731183e-06,
|
| 217 |
+
"loss": 0.6911,
|
| 218 |
+
"step": 7
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.93,
|
| 222 |
+
"logps_train/chosen": -106.92861938476562,
|
| 223 |
+
"logps_train/ref_chosen": -107.0,
|
| 224 |
+
"logps_train/ref_rejected": -104.0,
|
| 225 |
+
"logps_train/rejected": -104.0421142578125,
|
| 226 |
+
"rewards_train/accuracies": 0.859375,
|
| 227 |
+
"rewards_train/chosen": 0.18815864622592926,
|
| 228 |
+
"rewards_train/margins": 0.27952703833580017,
|
| 229 |
+
"rewards_train/rejected": -0.09136839210987091,
|
| 230 |
+
"step": 7
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.93,
|
| 234 |
+
"logps_train/chosen": -95.47774505615234,
|
| 235 |
+
"logps_train/ref_chosen": -96.0,
|
| 236 |
+
"logps_train/ref_rejected": -90.5,
|
| 237 |
+
"logps_train/rejected": -90.31706237792969,
|
| 238 |
+
"rewards_train/accuracies": 0.75,
|
| 239 |
+
"rewards_train/chosen": 0.20924821496009827,
|
| 240 |
+
"rewards_train/margins": 0.21952810510993004,
|
| 241 |
+
"rewards_train/rejected": -0.010279890149831772,
|
| 242 |
+
"step": 7
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 1.07,
|
| 246 |
+
"learning_rate": 3.634976249348867e-06,
|
| 247 |
+
"loss": 0.5866,
|
| 248 |
+
"step": 8
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 1.07,
|
| 252 |
+
"logps_train/chosen": -103.60890197753906,
|
| 253 |
+
"logps_train/ref_chosen": -104.0,
|
| 254 |
+
"logps_train/ref_rejected": -91.0,
|
| 255 |
+
"logps_train/rejected": -91.19517517089844,
|
| 256 |
+
"rewards_train/accuracies": 0.78125,
|
| 257 |
+
"rewards_train/chosen": 0.22508788108825684,
|
| 258 |
+
"rewards_train/margins": 0.2580987438559532,
|
| 259 |
+
"rewards_train/rejected": -0.03301086276769638,
|
| 260 |
+
"step": 8
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 1.07,
|
| 264 |
+
"logps_train/chosen": -89.79802703857422,
|
| 265 |
+
"logps_train/ref_chosen": -90.0,
|
| 266 |
+
"logps_train/ref_rejected": -91.0,
|
| 267 |
+
"logps_train/rejected": -91.28385925292969,
|
| 268 |
+
"rewards_train/accuracies": 0.796875,
|
| 269 |
+
"rewards_train/chosen": 0.18155044317245483,
|
| 270 |
+
"rewards_train/margins": 0.24022718518972397,
|
| 271 |
+
"rewards_train/rejected": -0.058676742017269135,
|
| 272 |
+
"step": 8
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 1.2,
|
| 276 |
+
"learning_rate": 3.272542485937369e-06,
|
| 277 |
+
"loss": 0.5875,
|
| 278 |
+
"step": 9
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 1.2,
|
| 282 |
+
"logps_train/chosen": -99.75357818603516,
|
| 283 |
+
"logps_train/ref_chosen": -100.0,
|
| 284 |
+
"logps_train/ref_rejected": -91.5,
|
| 285 |
+
"logps_train/rejected": -91.54853820800781,
|
| 286 |
+
"rewards_train/accuracies": 0.765625,
|
| 287 |
+
"rewards_train/chosen": 0.2291664034128189,
|
| 288 |
+
"rewards_train/margins": 0.28212443739175797,
|
| 289 |
+
"rewards_train/rejected": -0.052958033978939056,
|
| 290 |
+
"step": 9
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 1.2,
|
| 294 |
+
"logps_train/chosen": -97.27603149414062,
|
| 295 |
+
"logps_train/ref_chosen": -98.0,
|
| 296 |
+
"logps_train/ref_rejected": -95.5,
|
| 297 |
+
"logps_train/rejected": -95.36297607421875,
|
| 298 |
+
"rewards_train/accuracies": 0.828125,
|
| 299 |
+
"rewards_train/chosen": 0.23210078477859497,
|
| 300 |
+
"rewards_train/margins": 0.23072939738631248,
|
| 301 |
+
"rewards_train/rejected": 0.001371387392282486,
|
| 302 |
+
"step": 9
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 1.33,
|
| 306 |
+
"learning_rate": 2.8910861626005774e-06,
|
| 307 |
+
"loss": 0.583,
|
| 308 |
+
"step": 10
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 1.33,
|
| 312 |
+
"logps_train/chosen": -99.11138916015625,
|
| 313 |
+
"logps_train/ref_chosen": -99.5,
|
| 314 |
+
"logps_train/ref_rejected": -97.0,
|
| 315 |
+
"logps_train/rejected": -96.76425170898438,
|
| 316 |
+
"rewards_train/accuracies": 0.765625,
|
| 317 |
+
"rewards_train/chosen": 0.24630612134933472,
|
| 318 |
+
"rewards_train/margins": 0.23292576521635056,
|
| 319 |
+
"rewards_train/rejected": 0.013380356132984161,
|
| 320 |
+
"step": 10
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 1.33,
|
| 324 |
+
"logps_train/chosen": -90.8442611694336,
|
| 325 |
+
"logps_train/ref_chosen": -91.0,
|
| 326 |
+
"logps_train/ref_rejected": -97.5,
|
| 327 |
+
"logps_train/rejected": -97.66960144042969,
|
| 328 |
+
"rewards_train/accuracies": 0.765625,
|
| 329 |
+
"rewards_train/chosen": 0.26005905866622925,
|
| 330 |
+
"rewards_train/margins": 0.2439107522368431,
|
| 331 |
+
"rewards_train/rejected": 0.01614830642938614,
|
| 332 |
+
"step": 10
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 1.47,
|
| 336 |
+
"learning_rate": 2.5e-06,
|
| 337 |
+
"loss": 0.5913,
|
| 338 |
+
"step": 11
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.47,
|
| 342 |
+
"logps_train/chosen": -96.24613952636719,
|
| 343 |
+
"logps_train/ref_chosen": -97.0,
|
| 344 |
+
"logps_train/ref_rejected": -99.0,
|
| 345 |
+
"logps_train/rejected": -98.8271484375,
|
| 346 |
+
"rewards_train/accuracies": 0.84375,
|
| 347 |
+
"rewards_train/chosen": 0.23923271894454956,
|
| 348 |
+
"rewards_train/margins": 0.3027096092700958,
|
| 349 |
+
"rewards_train/rejected": -0.06347689032554626,
|
| 350 |
+
"step": 11
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 1.47,
|
| 354 |
+
"logps_train/chosen": -80.69171142578125,
|
| 355 |
+
"logps_train/ref_chosen": -81.0,
|
| 356 |
+
"logps_train/ref_rejected": -90.0,
|
| 357 |
+
"logps_train/rejected": -90.06707000732422,
|
| 358 |
+
"rewards_train/accuracies": 0.796875,
|
| 359 |
+
"rewards_train/chosen": 0.21196044981479645,
|
| 360 |
+
"rewards_train/margins": 0.21380476094782352,
|
| 361 |
+
"rewards_train/rejected": -0.0018443111330270767,
|
| 362 |
+
"step": 11
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 1.6,
|
| 366 |
+
"learning_rate": 2.1089138373994226e-06,
|
| 367 |
+
"loss": 0.5821,
|
| 368 |
+
"step": 12
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 1.6,
|
| 372 |
+
"logps_train/chosen": -89.44386291503906,
|
| 373 |
+
"logps_train/ref_chosen": -90.0,
|
| 374 |
+
"logps_train/ref_rejected": -96.0,
|
| 375 |
+
"logps_train/rejected": -95.8780746459961,
|
| 376 |
+
"rewards_train/accuracies": 0.796875,
|
| 377 |
+
"rewards_train/chosen": 0.2612239420413971,
|
| 378 |
+
"rewards_train/margins": 0.28380540758371353,
|
| 379 |
+
"rewards_train/rejected": -0.022581465542316437,
|
| 380 |
+
"step": 12
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.6,
|
| 384 |
+
"logps_train/chosen": -82.67738342285156,
|
| 385 |
+
"logps_train/ref_chosen": -83.0,
|
| 386 |
+
"logps_train/ref_rejected": -85.0,
|
| 387 |
+
"logps_train/rejected": -85.28672790527344,
|
| 388 |
+
"rewards_train/accuracies": 0.859375,
|
| 389 |
+
"rewards_train/chosen": 0.2267371118068695,
|
| 390 |
+
"rewards_train/margins": 0.2917319983243942,
|
| 391 |
+
"rewards_train/rejected": -0.06499488651752472,
|
| 392 |
+
"step": 12
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 1.73,
|
| 396 |
+
"learning_rate": 1.7274575140626318e-06,
|
| 397 |
+
"loss": 0.5736,
|
| 398 |
+
"step": 13
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 1.73,
|
| 402 |
+
"logps_train/chosen": -104.39497375488281,
|
| 403 |
+
"logps_train/ref_chosen": -105.0,
|
| 404 |
+
"logps_train/ref_rejected": -101.0,
|
| 405 |
+
"logps_train/rejected": -101.02708435058594,
|
| 406 |
+
"rewards_train/accuracies": 0.921875,
|
| 407 |
+
"rewards_train/chosen": 0.34890174865722656,
|
| 408 |
+
"rewards_train/margins": 0.4168848916888237,
|
| 409 |
+
"rewards_train/rejected": -0.06798314303159714,
|
| 410 |
+
"step": 13
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 1.73,
|
| 414 |
+
"logps_train/chosen": -89.95936584472656,
|
| 415 |
+
"logps_train/ref_chosen": -91.0,
|
| 416 |
+
"logps_train/ref_rejected": -85.0,
|
| 417 |
+
"logps_train/rejected": -85.16304016113281,
|
| 418 |
+
"rewards_train/accuracies": 0.875,
|
| 419 |
+
"rewards_train/chosen": 0.283745139837265,
|
| 420 |
+
"rewards_train/margins": 0.3208296597003937,
|
| 421 |
+
"rewards_train/rejected": -0.03708451986312866,
|
| 422 |
+
"step": 13
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.87,
|
| 426 |
+
"learning_rate": 1.3650237506511333e-06,
|
| 427 |
+
"loss": 0.5417,
|
| 428 |
+
"step": 14
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 1.87,
|
| 432 |
+
"logps_train/chosen": -92.31253051757812,
|
| 433 |
+
"logps_train/ref_chosen": -93.0,
|
| 434 |
+
"logps_train/ref_rejected": -108.5,
|
| 435 |
+
"logps_train/rejected": -108.66498565673828,
|
| 436 |
+
"rewards_train/accuracies": 0.796875,
|
| 437 |
+
"rewards_train/chosen": 0.2648744583129883,
|
| 438 |
+
"rewards_train/margins": 0.3167903497815132,
|
| 439 |
+
"rewards_train/rejected": -0.05191589146852493,
|
| 440 |
+
"step": 14
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 1.87,
|
| 444 |
+
"logps_train/chosen": -93.20501708984375,
|
| 445 |
+
"logps_train/ref_chosen": -94.0,
|
| 446 |
+
"logps_train/ref_rejected": -93.0,
|
| 447 |
+
"logps_train/rejected": -93.12667846679688,
|
| 448 |
+
"rewards_train/accuracies": 0.890625,
|
| 449 |
+
"rewards_train/chosen": 0.41396820545196533,
|
| 450 |
+
"rewards_train/margins": 0.4813934415578842,
|
| 451 |
+
"rewards_train/rejected": -0.06742523610591888,
|
| 452 |
+
"step": 14
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 2.0,
|
| 456 |
+
"learning_rate": 1.0305368692688175e-06,
|
| 457 |
+
"loss": 0.5331,
|
| 458 |
+
"step": 15
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 2.0,
|
| 462 |
+
"logps_train/chosen": -96.6799087524414,
|
| 463 |
+
"logps_train/ref_chosen": -97.5,
|
| 464 |
+
"logps_train/ref_rejected": -96.5,
|
| 465 |
+
"logps_train/rejected": -96.50393676757812,
|
| 466 |
+
"rewards_train/accuracies": 0.9375,
|
| 467 |
+
"rewards_train/chosen": 0.3427841067314148,
|
| 468 |
+
"rewards_train/margins": 0.3952302448451519,
|
| 469 |
+
"rewards_train/rejected": -0.052446138113737106,
|
| 470 |
+
"step": 15
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 2.0,
|
| 474 |
+
"logps_train/chosen": -92.8008041381836,
|
| 475 |
+
"logps_train/ref_chosen": -93.5,
|
| 476 |
+
"logps_train/ref_rejected": -106.0,
|
| 477 |
+
"logps_train/rejected": -105.96940612792969,
|
| 478 |
+
"rewards_train/accuracies": 0.828125,
|
| 479 |
+
"rewards_train/chosen": 0.3604634702205658,
|
| 480 |
+
"rewards_train/margins": 0.4462389573454857,
|
| 481 |
+
"rewards_train/rejected": -0.08577548712491989,
|
| 482 |
+
"step": 15
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 2.13,
|
| 486 |
+
"learning_rate": 7.322330470336314e-07,
|
| 487 |
+
"loss": 0.5232,
|
| 488 |
+
"step": 16
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 2.13,
|
| 492 |
+
"logps_train/chosen": -94.60572814941406,
|
| 493 |
+
"logps_train/ref_chosen": -95.0,
|
| 494 |
+
"logps_train/ref_rejected": -89.0,
|
| 495 |
+
"logps_train/rejected": -89.16722869873047,
|
| 496 |
+
"rewards_train/accuracies": 0.953125,
|
| 497 |
+
"rewards_train/chosen": 0.3311697542667389,
|
| 498 |
+
"rewards_train/margins": 0.4457896798849106,
|
| 499 |
+
"rewards_train/rejected": -0.11461992561817169,
|
| 500 |
+
"step": 16
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 2.13,
|
| 504 |
+
"logps_train/chosen": -93.5136489868164,
|
| 505 |
+
"logps_train/ref_chosen": -94.0,
|
| 506 |
+
"logps_train/ref_rejected": -88.5,
|
| 507 |
+
"logps_train/rejected": -88.58706665039062,
|
| 508 |
+
"rewards_train/accuracies": 0.875,
|
| 509 |
+
"rewards_train/chosen": 0.31861382722854614,
|
| 510 |
+
"rewards_train/margins": 0.35628755390644073,
|
| 511 |
+
"rewards_train/rejected": -0.03767372667789459,
|
| 512 |
+
"step": 16
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 2.27,
|
| 516 |
+
"learning_rate": 4.774575140626317e-07,
|
| 517 |
+
"loss": 0.5274,
|
| 518 |
+
"step": 17
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 2.27,
|
| 522 |
+
"logps_train/chosen": -95.30484771728516,
|
| 523 |
+
"logps_train/ref_chosen": -96.0,
|
| 524 |
+
"logps_train/ref_rejected": -96.0,
|
| 525 |
+
"logps_train/rejected": -96.07192993164062,
|
| 526 |
+
"rewards_train/accuracies": 0.921875,
|
| 527 |
+
"rewards_train/chosen": 0.3683273196220398,
|
| 528 |
+
"rewards_train/margins": 0.48193351179361343,
|
| 529 |
+
"rewards_train/rejected": -0.11360619217157364,
|
| 530 |
+
"step": 17
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 2.27,
|
| 534 |
+
"logps_train/chosen": -90.89393615722656,
|
| 535 |
+
"logps_train/ref_chosen": -91.5,
|
| 536 |
+
"logps_train/ref_rejected": -100.0,
|
| 537 |
+
"logps_train/rejected": -100.20901489257812,
|
| 538 |
+
"rewards_train/accuracies": 0.9375,
|
| 539 |
+
"rewards_train/chosen": 0.34770655632019043,
|
| 540 |
+
"rewards_train/margins": 0.4666781648993492,
|
| 541 |
+
"rewards_train/rejected": -0.11897160857915878,
|
| 542 |
+
"step": 17
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 2.4,
|
| 546 |
+
"learning_rate": 2.7248368952908055e-07,
|
| 547 |
+
"loss": 0.4996,
|
| 548 |
+
"step": 18
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 2.4,
|
| 552 |
+
"logps_train/chosen": -101.5006103515625,
|
| 553 |
+
"logps_train/ref_chosen": -102.0,
|
| 554 |
+
"logps_train/ref_rejected": -99.0,
|
| 555 |
+
"logps_train/rejected": -98.9917221069336,
|
| 556 |
+
"rewards_train/accuracies": 0.921875,
|
| 557 |
+
"rewards_train/chosen": 0.39098966121673584,
|
| 558 |
+
"rewards_train/margins": 0.46887996047735214,
|
| 559 |
+
"rewards_train/rejected": -0.0778902992606163,
|
| 560 |
+
"step": 18
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 2.4,
|
| 564 |
+
"logps_train/chosen": -100.89315795898438,
|
| 565 |
+
"logps_train/ref_chosen": -101.5,
|
| 566 |
+
"logps_train/ref_rejected": -101.0,
|
| 567 |
+
"logps_train/rejected": -101.20696258544922,
|
| 568 |
+
"rewards_train/accuracies": 0.875,
|
| 569 |
+
"rewards_train/chosen": 0.3883822560310364,
|
| 570 |
+
"rewards_train/margins": 0.4976026713848114,
|
| 571 |
+
"rewards_train/rejected": -0.10922041535377502,
|
| 572 |
+
"step": 18
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 2.53,
|
| 576 |
+
"learning_rate": 1.223587092621162e-07,
|
| 577 |
+
"loss": 0.499,
|
| 578 |
+
"step": 19
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 2.53,
|
| 582 |
+
"logps_train/chosen": -93.89166259765625,
|
| 583 |
+
"logps_train/ref_chosen": -94.5,
|
| 584 |
+
"logps_train/ref_rejected": -101.0,
|
| 585 |
+
"logps_train/rejected": -101.55459594726562,
|
| 586 |
+
"rewards_train/accuracies": 0.859375,
|
| 587 |
+
"rewards_train/chosen": 0.33444538712501526,
|
| 588 |
+
"rewards_train/margins": 0.45895563066005707,
|
| 589 |
+
"rewards_train/rejected": -0.12451024353504181,
|
| 590 |
+
"step": 19
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 2.53,
|
| 594 |
+
"logps_train/chosen": -100.19595336914062,
|
| 595 |
+
"logps_train/ref_chosen": -101.0,
|
| 596 |
+
"logps_train/ref_rejected": -95.5,
|
| 597 |
+
"logps_train/rejected": -95.7244873046875,
|
| 598 |
+
"rewards_train/accuracies": 0.921875,
|
| 599 |
+
"rewards_train/chosen": 0.38786405324935913,
|
| 600 |
+
"rewards_train/margins": 0.44200433045625687,
|
| 601 |
+
"rewards_train/rejected": -0.054140277206897736,
|
| 602 |
+
"step": 19
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 2.67,
|
| 606 |
+
"learning_rate": 3.077914851215585e-08,
|
| 607 |
+
"loss": 0.5114,
|
| 608 |
+
"step": 20
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 2.67,
|
| 612 |
+
"logps_train/chosen": -75.63420867919922,
|
| 613 |
+
"logps_train/ref_chosen": -76.0,
|
| 614 |
+
"logps_train/ref_rejected": -78.0,
|
| 615 |
+
"logps_train/rejected": -78.19580841064453,
|
| 616 |
+
"rewards_train/accuracies": 0.828125,
|
| 617 |
+
"rewards_train/chosen": 0.28807422518730164,
|
| 618 |
+
"rewards_train/margins": 0.3581909090280533,
|
| 619 |
+
"rewards_train/rejected": -0.07011668384075165,
|
| 620 |
+
"step": 20
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 2.67,
|
| 624 |
+
"logps_train/chosen": -94.97181701660156,
|
| 625 |
+
"logps_train/ref_chosen": -96.0,
|
| 626 |
+
"logps_train/ref_rejected": -94.0,
|
| 627 |
+
"logps_train/rejected": -94.27290344238281,
|
| 628 |
+
"rewards_train/accuracies": 0.921875,
|
| 629 |
+
"rewards_train/chosen": 0.38176849484443665,
|
| 630 |
+
"rewards_train/margins": 0.46939007192850113,
|
| 631 |
+
"rewards_train/rejected": -0.08762157708406448,
|
| 632 |
+
"step": 20
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 2.8,
|
| 636 |
+
"learning_rate": 0.0,
|
| 637 |
+
"loss": 0.5237,
|
| 638 |
+
"step": 21
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 2.8,
|
| 642 |
+
"step": 21,
|
| 643 |
+
"total_flos": 0.0,
|
| 644 |
+
"train_loss": 0.5982543017183032,
|
| 645 |
+
"train_runtime": 243.8656,
|
| 646 |
+
"train_samples_per_second": 11.182,
|
| 647 |
+
"train_steps_per_second": 0.086
|
| 648 |
+
}
|
| 649 |
+
],
|
| 650 |
+
"max_steps": 21,
|
| 651 |
+
"num_train_epochs": 3,
|
| 652 |
+
"total_flos": 0.0,
|
| 653 |
+
"trial_name": null,
|
| 654 |
+
"trial_params": null
|
| 655 |
+
}
|