Commit ·
3d06d19
1
Parent(s): 33502f4
add privacy-tuned model
Browse files
tinyllava_demo/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
|
tinyllava_demo/adapter_config.json
ADDED
|
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"auto_mapping": null,
|
| 3 |
+
"base_model_name_or_path": "bczhou/TinyLLaVA-3.1B",
|
| 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": 64,
|
| 11 |
+
"lora_dropout": 0.05,
|
| 12 |
+
"modules_to_save": null,
|
| 13 |
+
"peft_type": "LORA",
|
| 14 |
+
"r": 32,
|
| 15 |
+
"revision": null,
|
| 16 |
+
"target_modules": [
|
| 17 |
+
"fc2",
|
| 18 |
+
"dense",
|
| 19 |
+
"fc1",
|
| 20 |
+
"k_proj",
|
| 21 |
+
"q_proj",
|
| 22 |
+
"v_proj"
|
| 23 |
+
],
|
| 24 |
+
"task_type": "CAUSAL_LM"
|
| 25 |
+
}
|
tinyllava_demo/adapter_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:38423a6fa4f3ea259000fcdd959b64359899a7fd3d0622104ad8d0b4885edbb5
|
| 3 |
+
size 94504269
|
tinyllava_demo/config.json
ADDED
|
@@ -0,0 +1,54 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "bczhou/TinyLLaVA-3.1B",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"TinyLlavaPhiForCausalLM"
|
| 5 |
+
],
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"auto_map": {
|
| 8 |
+
"AutoConfig": "microsoft/phi-2--configuration_phi.PhiConfig",
|
| 9 |
+
"AutoModelForCausalLM": "microsoft/phi-2--modeling_phi.PhiForCausalLM"
|
| 10 |
+
},
|
| 11 |
+
"bos_token_id": 50256,
|
| 12 |
+
"embd_pdrop": 0.0,
|
| 13 |
+
"eos_token_id": 50256,
|
| 14 |
+
"freeze_mm_mlp_adapter": false,
|
| 15 |
+
"hidden_act": "gelu_new",
|
| 16 |
+
"hidden_size": 2560,
|
| 17 |
+
"image_aspect_ratio": "pad",
|
| 18 |
+
"initializer_range": 0.02,
|
| 19 |
+
"intermediate_size": 10240,
|
| 20 |
+
"layer_norm_eps": 1e-05,
|
| 21 |
+
"max_position_embeddings": 2048,
|
| 22 |
+
"mm_hidden_size": 1152,
|
| 23 |
+
"mm_patch_merge_type": "flat",
|
| 24 |
+
"mm_projector_lr": null,
|
| 25 |
+
"mm_projector_type": "mlp2x_gelu",
|
| 26 |
+
"mm_use_im_patch_token": false,
|
| 27 |
+
"mm_use_im_start_end": false,
|
| 28 |
+
"mm_vision_select_feature": "patch",
|
| 29 |
+
"mm_vision_select_layer": -2,
|
| 30 |
+
"mm_vision_tower": "bczhou/TinyLLaVA-3.1B-SigLIP",
|
| 31 |
+
"model_type": "tiny_llava_phi",
|
| 32 |
+
"num_attention_heads": 32,
|
| 33 |
+
"num_hidden_layers": 32,
|
| 34 |
+
"num_key_value_heads": 32,
|
| 35 |
+
"pad_token": "<|endoftext|>",
|
| 36 |
+
"pad_token_id": 50256,
|
| 37 |
+
"partial_rotary_factor": 0.4,
|
| 38 |
+
"qk_layernorm": false,
|
| 39 |
+
"resid_pdrop": 0.1,
|
| 40 |
+
"rope_scaling": null,
|
| 41 |
+
"rope_theta": 10000.0,
|
| 42 |
+
"tie_word_embeddings": false,
|
| 43 |
+
"tokenizer_model_max_length": 3072,
|
| 44 |
+
"tokenizer_padding_side": "right",
|
| 45 |
+
"torch_dtype": "float16",
|
| 46 |
+
"transformers_version": "4.37.2",
|
| 47 |
+
"tune_embed_tokens": false,
|
| 48 |
+
"tune_entire_model": false,
|
| 49 |
+
"tune_mm_mlp_adapter": false,
|
| 50 |
+
"tune_vision_tower": false,
|
| 51 |
+
"use_cache": true,
|
| 52 |
+
"use_mm_proj": true,
|
| 53 |
+
"vocab_size": 51200
|
| 54 |
+
}
|
tinyllava_demo/non_lora_trainables.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:62da77dc2703278c5776208e19caf191380873e73d72b60c0c2b42e33464d767
|
| 3 |
+
size 19017447
|
tinyllava_demo/trainer_state.json
ADDED
|
@@ -0,0 +1,1230 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 20.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 200,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.1,
|
| 13 |
+
"learning_rate": 3.3333333333333333e-06,
|
| 14 |
+
"loss": 1.6584,
|
| 15 |
+
"step": 1
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 0.2,
|
| 19 |
+
"learning_rate": 6.666666666666667e-06,
|
| 20 |
+
"loss": 1.6844,
|
| 21 |
+
"step": 2
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.3,
|
| 25 |
+
"learning_rate": 1e-05,
|
| 26 |
+
"loss": 1.6431,
|
| 27 |
+
"step": 3
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 0.4,
|
| 31 |
+
"learning_rate": 1.3333333333333333e-05,
|
| 32 |
+
"loss": 1.6374,
|
| 33 |
+
"step": 4
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 0.5,
|
| 37 |
+
"learning_rate": 1.6666666666666667e-05,
|
| 38 |
+
"loss": 1.7387,
|
| 39 |
+
"step": 5
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.6,
|
| 43 |
+
"learning_rate": 2e-05,
|
| 44 |
+
"loss": 1.6855,
|
| 45 |
+
"step": 6
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.7,
|
| 49 |
+
"learning_rate": 1.9998688836656322e-05,
|
| 50 |
+
"loss": 1.6148,
|
| 51 |
+
"step": 7
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.8,
|
| 55 |
+
"learning_rate": 1.9994755690455154e-05,
|
| 56 |
+
"loss": 1.6946,
|
| 57 |
+
"step": 8
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.9,
|
| 61 |
+
"learning_rate": 1.998820159279591e-05,
|
| 62 |
+
"loss": 1.7163,
|
| 63 |
+
"step": 9
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 1.0,
|
| 67 |
+
"learning_rate": 1.997902826237712e-05,
|
| 68 |
+
"loss": 1.6442,
|
| 69 |
+
"step": 10
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 1.1,
|
| 73 |
+
"learning_rate": 1.9967238104745695e-05,
|
| 74 |
+
"loss": 1.5848,
|
| 75 |
+
"step": 11
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 1.2,
|
| 79 |
+
"learning_rate": 1.995283421166614e-05,
|
| 80 |
+
"loss": 1.5664,
|
| 81 |
+
"step": 12
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 1.3,
|
| 85 |
+
"learning_rate": 1.993582036030978e-05,
|
| 86 |
+
"loss": 1.6251,
|
| 87 |
+
"step": 13
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 1.4,
|
| 91 |
+
"learning_rate": 1.9916201012264255e-05,
|
| 92 |
+
"loss": 1.5254,
|
| 93 |
+
"step": 14
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 1.5,
|
| 97 |
+
"learning_rate": 1.9893981312363563e-05,
|
| 98 |
+
"loss": 1.5047,
|
| 99 |
+
"step": 15
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 1.6,
|
| 103 |
+
"learning_rate": 1.9869167087338908e-05,
|
| 104 |
+
"loss": 1.475,
|
| 105 |
+
"step": 16
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 1.7,
|
| 109 |
+
"learning_rate": 1.9841764844290744e-05,
|
| 110 |
+
"loss": 1.4482,
|
| 111 |
+
"step": 17
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 1.8,
|
| 115 |
+
"learning_rate": 1.9811781768982392e-05,
|
| 116 |
+
"loss": 1.4903,
|
| 117 |
+
"step": 18
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 1.9,
|
| 121 |
+
"learning_rate": 1.977922572395571e-05,
|
| 122 |
+
"loss": 1.4799,
|
| 123 |
+
"step": 19
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 2.0,
|
| 127 |
+
"learning_rate": 1.9744105246469264e-05,
|
| 128 |
+
"loss": 1.3944,
|
| 129 |
+
"step": 20
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 2.1,
|
| 133 |
+
"learning_rate": 1.9706429546259592e-05,
|
| 134 |
+
"loss": 1.3967,
|
| 135 |
+
"step": 21
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 2.2,
|
| 139 |
+
"learning_rate": 1.9666208503126115e-05,
|
| 140 |
+
"loss": 1.389,
|
| 141 |
+
"step": 22
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 2.3,
|
| 145 |
+
"learning_rate": 1.9623452664340305e-05,
|
| 146 |
+
"loss": 1.3755,
|
| 147 |
+
"step": 23
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 2.4,
|
| 151 |
+
"learning_rate": 1.957817324187987e-05,
|
| 152 |
+
"loss": 1.3328,
|
| 153 |
+
"step": 24
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 2.5,
|
| 157 |
+
"learning_rate": 1.953038210948861e-05,
|
| 158 |
+
"loss": 1.3103,
|
| 159 |
+
"step": 25
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 2.6,
|
| 163 |
+
"learning_rate": 1.9480091799562706e-05,
|
| 164 |
+
"loss": 1.2674,
|
| 165 |
+
"step": 26
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 2.7,
|
| 169 |
+
"learning_rate": 1.9427315499864345e-05,
|
| 170 |
+
"loss": 1.2472,
|
| 171 |
+
"step": 27
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 2.8,
|
| 175 |
+
"learning_rate": 1.937206705006344e-05,
|
| 176 |
+
"loss": 1.3056,
|
| 177 |
+
"step": 28
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 2.9,
|
| 181 |
+
"learning_rate": 1.9314360938108427e-05,
|
| 182 |
+
"loss": 1.3436,
|
| 183 |
+
"step": 29
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 3.0,
|
| 187 |
+
"learning_rate": 1.9254212296427043e-05,
|
| 188 |
+
"loss": 1.2784,
|
| 189 |
+
"step": 30
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 3.1,
|
| 193 |
+
"learning_rate": 1.9191636897958123e-05,
|
| 194 |
+
"loss": 1.2273,
|
| 195 |
+
"step": 31
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 3.2,
|
| 199 |
+
"learning_rate": 1.9126651152015404e-05,
|
| 200 |
+
"loss": 1.2456,
|
| 201 |
+
"step": 32
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 3.3,
|
| 205 |
+
"learning_rate": 1.905927209998447e-05,
|
| 206 |
+
"loss": 1.2321,
|
| 207 |
+
"step": 33
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 3.4,
|
| 211 |
+
"learning_rate": 1.8989517410853956e-05,
|
| 212 |
+
"loss": 1.187,
|
| 213 |
+
"step": 34
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 3.5,
|
| 217 |
+
"learning_rate": 1.8917405376582144e-05,
|
| 218 |
+
"loss": 1.1412,
|
| 219 |
+
"step": 35
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 3.6,
|
| 223 |
+
"learning_rate": 1.8842954907300236e-05,
|
| 224 |
+
"loss": 1.1938,
|
| 225 |
+
"step": 36
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 3.7,
|
| 229 |
+
"learning_rate": 1.876618552635348e-05,
|
| 230 |
+
"loss": 1.2093,
|
| 231 |
+
"step": 37
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 3.8,
|
| 235 |
+
"learning_rate": 1.8687117365181514e-05,
|
| 236 |
+
"loss": 1.1224,
|
| 237 |
+
"step": 38
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 3.9,
|
| 241 |
+
"learning_rate": 1.8605771158039253e-05,
|
| 242 |
+
"loss": 1.1887,
|
| 243 |
+
"step": 39
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 4.0,
|
| 247 |
+
"learning_rate": 1.8522168236559693e-05,
|
| 248 |
+
"loss": 1.1597,
|
| 249 |
+
"step": 40
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 4.1,
|
| 253 |
+
"learning_rate": 1.8436330524160048e-05,
|
| 254 |
+
"loss": 1.1533,
|
| 255 |
+
"step": 41
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 4.2,
|
| 259 |
+
"learning_rate": 1.8348280530292712e-05,
|
| 260 |
+
"loss": 1.1125,
|
| 261 |
+
"step": 42
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 4.3,
|
| 265 |
+
"learning_rate": 1.8258041344542567e-05,
|
| 266 |
+
"loss": 1.1465,
|
| 267 |
+
"step": 43
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 4.4,
|
| 271 |
+
"learning_rate": 1.816563663057211e-05,
|
| 272 |
+
"loss": 1.126,
|
| 273 |
+
"step": 44
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 4.5,
|
| 277 |
+
"learning_rate": 1.8071090619916095e-05,
|
| 278 |
+
"loss": 1.0512,
|
| 279 |
+
"step": 45
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 4.6,
|
| 283 |
+
"learning_rate": 1.797442810562721e-05,
|
| 284 |
+
"loss": 1.0602,
|
| 285 |
+
"step": 46
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 4.7,
|
| 289 |
+
"learning_rate": 1.7875674435774546e-05,
|
| 290 |
+
"loss": 1.0927,
|
| 291 |
+
"step": 47
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 4.8,
|
| 295 |
+
"learning_rate": 1.7774855506796497e-05,
|
| 296 |
+
"loss": 1.109,
|
| 297 |
+
"step": 48
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 4.9,
|
| 301 |
+
"learning_rate": 1.767199775670986e-05,
|
| 302 |
+
"loss": 1.0789,
|
| 303 |
+
"step": 49
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 5.0,
|
| 307 |
+
"learning_rate": 1.7567128158176955e-05,
|
| 308 |
+
"loss": 1.0283,
|
| 309 |
+
"step": 50
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 5.1,
|
| 313 |
+
"learning_rate": 1.7460274211432463e-05,
|
| 314 |
+
"loss": 0.9957,
|
| 315 |
+
"step": 51
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 5.2,
|
| 319 |
+
"learning_rate": 1.7351463937072008e-05,
|
| 320 |
+
"loss": 1.0582,
|
| 321 |
+
"step": 52
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 5.3,
|
| 325 |
+
"learning_rate": 1.7240725868704218e-05,
|
| 326 |
+
"loss": 1.0579,
|
| 327 |
+
"step": 53
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 5.4,
|
| 331 |
+
"learning_rate": 1.7128089045468294e-05,
|
| 332 |
+
"loss": 1.042,
|
| 333 |
+
"step": 54
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 5.5,
|
| 337 |
+
"learning_rate": 1.7013583004418994e-05,
|
| 338 |
+
"loss": 1.0655,
|
| 339 |
+
"step": 55
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 5.6,
|
| 343 |
+
"learning_rate": 1.6897237772781046e-05,
|
| 344 |
+
"loss": 0.9831,
|
| 345 |
+
"step": 56
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 5.7,
|
| 349 |
+
"learning_rate": 1.6779083860075032e-05,
|
| 350 |
+
"loss": 1.0056,
|
| 351 |
+
"step": 57
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 5.8,
|
| 355 |
+
"learning_rate": 1.665915225011681e-05,
|
| 356 |
+
"loss": 1.0325,
|
| 357 |
+
"step": 58
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 5.9,
|
| 361 |
+
"learning_rate": 1.6537474392892527e-05,
|
| 362 |
+
"loss": 1.0187,
|
| 363 |
+
"step": 59
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 6.0,
|
| 367 |
+
"learning_rate": 1.6414082196311402e-05,
|
| 368 |
+
"loss": 0.9638,
|
| 369 |
+
"step": 60
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 6.1,
|
| 373 |
+
"learning_rate": 1.6289008017838447e-05,
|
| 374 |
+
"loss": 0.9851,
|
| 375 |
+
"step": 61
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 6.2,
|
| 379 |
+
"learning_rate": 1.6162284656009276e-05,
|
| 380 |
+
"loss": 0.9483,
|
| 381 |
+
"step": 62
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 6.3,
|
| 385 |
+
"learning_rate": 1.603394534182925e-05,
|
| 386 |
+
"loss": 0.9569,
|
| 387 |
+
"step": 63
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 6.4,
|
| 391 |
+
"learning_rate": 1.5904023730059227e-05,
|
| 392 |
+
"loss": 0.9335,
|
| 393 |
+
"step": 64
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 6.5,
|
| 397 |
+
"learning_rate": 1.5772553890390196e-05,
|
| 398 |
+
"loss": 0.9353,
|
| 399 |
+
"step": 65
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 6.6,
|
| 403 |
+
"learning_rate": 1.5639570298509067e-05,
|
| 404 |
+
"loss": 0.9462,
|
| 405 |
+
"step": 66
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 6.7,
|
| 409 |
+
"learning_rate": 1.5505107827058038e-05,
|
| 410 |
+
"loss": 0.9858,
|
| 411 |
+
"step": 67
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 6.8,
|
| 415 |
+
"learning_rate": 1.536920173648984e-05,
|
| 416 |
+
"loss": 0.9902,
|
| 417 |
+
"step": 68
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 6.9,
|
| 421 |
+
"learning_rate": 1.52318876658213e-05,
|
| 422 |
+
"loss": 0.9833,
|
| 423 |
+
"step": 69
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 7.0,
|
| 427 |
+
"learning_rate": 1.5093201623287631e-05,
|
| 428 |
+
"loss": 0.9841,
|
| 429 |
+
"step": 70
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 7.1,
|
| 433 |
+
"learning_rate": 1.4953179976899878e-05,
|
| 434 |
+
"loss": 0.9176,
|
| 435 |
+
"step": 71
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 7.2,
|
| 439 |
+
"learning_rate": 1.4811859444908053e-05,
|
| 440 |
+
"loss": 0.9687,
|
| 441 |
+
"step": 72
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 7.3,
|
| 445 |
+
"learning_rate": 1.4669277086172406e-05,
|
| 446 |
+
"loss": 0.9281,
|
| 447 |
+
"step": 73
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 7.4,
|
| 451 |
+
"learning_rate": 1.4525470290445392e-05,
|
| 452 |
+
"loss": 0.9099,
|
| 453 |
+
"step": 74
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 7.5,
|
| 457 |
+
"learning_rate": 1.4380476768566825e-05,
|
| 458 |
+
"loss": 0.9018,
|
| 459 |
+
"step": 75
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 7.6,
|
| 463 |
+
"learning_rate": 1.4234334542574906e-05,
|
| 464 |
+
"loss": 0.9051,
|
| 465 |
+
"step": 76
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 7.7,
|
| 469 |
+
"learning_rate": 1.4087081935735565e-05,
|
| 470 |
+
"loss": 0.9222,
|
| 471 |
+
"step": 77
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 7.8,
|
| 475 |
+
"learning_rate": 1.3938757562492873e-05,
|
| 476 |
+
"loss": 0.8984,
|
| 477 |
+
"step": 78
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 7.9,
|
| 481 |
+
"learning_rate": 1.378940031834307e-05,
|
| 482 |
+
"loss": 0.9621,
|
| 483 |
+
"step": 79
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 8.0,
|
| 487 |
+
"learning_rate": 1.3639049369634878e-05,
|
| 488 |
+
"loss": 0.8854,
|
| 489 |
+
"step": 80
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 8.1,
|
| 493 |
+
"learning_rate": 1.3487744143298822e-05,
|
| 494 |
+
"loss": 0.9116,
|
| 495 |
+
"step": 81
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 8.2,
|
| 499 |
+
"learning_rate": 1.3335524316508208e-05,
|
| 500 |
+
"loss": 0.9,
|
| 501 |
+
"step": 82
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 8.3,
|
| 505 |
+
"learning_rate": 1.3182429806274442e-05,
|
| 506 |
+
"loss": 0.8889,
|
| 507 |
+
"step": 83
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 8.4,
|
| 511 |
+
"learning_rate": 1.3028500758979507e-05,
|
| 512 |
+
"loss": 0.8858,
|
| 513 |
+
"step": 84
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 8.5,
|
| 517 |
+
"learning_rate": 1.2873777539848284e-05,
|
| 518 |
+
"loss": 0.8779,
|
| 519 |
+
"step": 85
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 8.6,
|
| 523 |
+
"learning_rate": 1.2718300722363431e-05,
|
| 524 |
+
"loss": 0.921,
|
| 525 |
+
"step": 86
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 8.7,
|
| 529 |
+
"learning_rate": 1.2562111077625723e-05,
|
| 530 |
+
"loss": 0.8331,
|
| 531 |
+
"step": 87
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 8.8,
|
| 535 |
+
"learning_rate": 1.2405249563662539e-05,
|
| 536 |
+
"loss": 0.85,
|
| 537 |
+
"step": 88
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 8.9,
|
| 541 |
+
"learning_rate": 1.2247757314687296e-05,
|
| 542 |
+
"loss": 0.8723,
|
| 543 |
+
"step": 89
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 9.0,
|
| 547 |
+
"learning_rate": 1.2089675630312755e-05,
|
| 548 |
+
"loss": 0.875,
|
| 549 |
+
"step": 90
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 9.1,
|
| 553 |
+
"learning_rate": 1.1931045964720882e-05,
|
| 554 |
+
"loss": 0.8863,
|
| 555 |
+
"step": 91
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 9.2,
|
| 559 |
+
"learning_rate": 1.177190991579223e-05,
|
| 560 |
+
"loss": 0.9104,
|
| 561 |
+
"step": 92
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 9.3,
|
| 565 |
+
"learning_rate": 1.1612309214197599e-05,
|
| 566 |
+
"loss": 0.8363,
|
| 567 |
+
"step": 93
|
| 568 |
+
},
|
| 569 |
+
{
|
| 570 |
+
"epoch": 9.4,
|
| 571 |
+
"learning_rate": 1.1452285712454905e-05,
|
| 572 |
+
"loss": 0.855,
|
| 573 |
+
"step": 94
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 9.5,
|
| 577 |
+
"learning_rate": 1.1291881373954066e-05,
|
| 578 |
+
"loss": 0.8453,
|
| 579 |
+
"step": 95
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 9.6,
|
| 583 |
+
"learning_rate": 1.1131138261952845e-05,
|
| 584 |
+
"loss": 0.8448,
|
| 585 |
+
"step": 96
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 9.7,
|
| 589 |
+
"learning_rate": 1.0970098528546482e-05,
|
| 590 |
+
"loss": 0.8287,
|
| 591 |
+
"step": 97
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 9.8,
|
| 595 |
+
"learning_rate": 1.0808804403614044e-05,
|
| 596 |
+
"loss": 0.861,
|
| 597 |
+
"step": 98
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 9.9,
|
| 601 |
+
"learning_rate": 1.0647298183744359e-05,
|
| 602 |
+
"loss": 0.8576,
|
| 603 |
+
"step": 99
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 10.0,
|
| 607 |
+
"learning_rate": 1.0485622221144485e-05,
|
| 608 |
+
"loss": 0.8056,
|
| 609 |
+
"step": 100
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 10.1,
|
| 613 |
+
"learning_rate": 1.0323818912533561e-05,
|
| 614 |
+
"loss": 0.871,
|
| 615 |
+
"step": 101
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 10.2,
|
| 619 |
+
"learning_rate": 1.0161930688025018e-05,
|
| 620 |
+
"loss": 0.8369,
|
| 621 |
+
"step": 102
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 10.3,
|
| 625 |
+
"learning_rate": 1e-05,
|
| 626 |
+
"loss": 0.7879,
|
| 627 |
+
"step": 103
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 10.4,
|
| 631 |
+
"learning_rate": 9.838069311974986e-06,
|
| 632 |
+
"loss": 0.8555,
|
| 633 |
+
"step": 104
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 10.5,
|
| 637 |
+
"learning_rate": 9.676181087466444e-06,
|
| 638 |
+
"loss": 0.8323,
|
| 639 |
+
"step": 105
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 10.6,
|
| 643 |
+
"learning_rate": 9.514377778855521e-06,
|
| 644 |
+
"loss": 0.8219,
|
| 645 |
+
"step": 106
|
| 646 |
+
},
|
| 647 |
+
{
|
| 648 |
+
"epoch": 10.7,
|
| 649 |
+
"learning_rate": 9.352701816255643e-06,
|
| 650 |
+
"loss": 0.8714,
|
| 651 |
+
"step": 107
|
| 652 |
+
},
|
| 653 |
+
{
|
| 654 |
+
"epoch": 10.8,
|
| 655 |
+
"learning_rate": 9.19119559638596e-06,
|
| 656 |
+
"loss": 0.8201,
|
| 657 |
+
"step": 108
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 10.9,
|
| 661 |
+
"learning_rate": 9.02990147145352e-06,
|
| 662 |
+
"loss": 0.8091,
|
| 663 |
+
"step": 109
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 11.0,
|
| 667 |
+
"learning_rate": 8.868861738047158e-06,
|
| 668 |
+
"loss": 0.7816,
|
| 669 |
+
"step": 110
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 11.1,
|
| 673 |
+
"learning_rate": 8.708118626045939e-06,
|
| 674 |
+
"loss": 0.7829,
|
| 675 |
+
"step": 111
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 11.2,
|
| 679 |
+
"learning_rate": 8.5477142875451e-06,
|
| 680 |
+
"loss": 0.7808,
|
| 681 |
+
"step": 112
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 11.3,
|
| 685 |
+
"learning_rate": 8.387690785802403e-06,
|
| 686 |
+
"loss": 0.8441,
|
| 687 |
+
"step": 113
|
| 688 |
+
},
|
| 689 |
+
{
|
| 690 |
+
"epoch": 11.4,
|
| 691 |
+
"learning_rate": 8.228090084207773e-06,
|
| 692 |
+
"loss": 0.8306,
|
| 693 |
+
"step": 114
|
| 694 |
+
},
|
| 695 |
+
{
|
| 696 |
+
"epoch": 11.5,
|
| 697 |
+
"learning_rate": 8.068954035279121e-06,
|
| 698 |
+
"loss": 0.7839,
|
| 699 |
+
"step": 115
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 11.6,
|
| 703 |
+
"learning_rate": 7.91032436968725e-06,
|
| 704 |
+
"loss": 0.8155,
|
| 705 |
+
"step": 116
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 11.7,
|
| 709 |
+
"learning_rate": 7.752242685312709e-06,
|
| 710 |
+
"loss": 0.7924,
|
| 711 |
+
"step": 117
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 11.8,
|
| 715 |
+
"learning_rate": 7.594750436337467e-06,
|
| 716 |
+
"loss": 0.8312,
|
| 717 |
+
"step": 118
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 11.9,
|
| 721 |
+
"learning_rate": 7.4378889223742766e-06,
|
| 722 |
+
"loss": 0.7709,
|
| 723 |
+
"step": 119
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 12.0,
|
| 727 |
+
"learning_rate": 7.2816992776365714e-06,
|
| 728 |
+
"loss": 0.8364,
|
| 729 |
+
"step": 120
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 12.1,
|
| 733 |
+
"learning_rate": 7.126222460151719e-06,
|
| 734 |
+
"loss": 0.7958,
|
| 735 |
+
"step": 121
|
| 736 |
+
},
|
| 737 |
+
{
|
| 738 |
+
"epoch": 12.2,
|
| 739 |
+
"learning_rate": 6.971499241020495e-06,
|
| 740 |
+
"loss": 0.8174,
|
| 741 |
+
"step": 122
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 12.3,
|
| 745 |
+
"learning_rate": 6.8175701937255645e-06,
|
| 746 |
+
"loss": 0.7507,
|
| 747 |
+
"step": 123
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 12.4,
|
| 751 |
+
"learning_rate": 6.664475683491797e-06,
|
| 752 |
+
"loss": 0.8045,
|
| 753 |
+
"step": 124
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 12.5,
|
| 757 |
+
"learning_rate": 6.5122558567011775e-06,
|
| 758 |
+
"loss": 0.7939,
|
| 759 |
+
"step": 125
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 12.6,
|
| 763 |
+
"learning_rate": 6.360950630365126e-06,
|
| 764 |
+
"loss": 0.7944,
|
| 765 |
+
"step": 126
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 12.7,
|
| 769 |
+
"learning_rate": 6.210599681656933e-06,
|
| 770 |
+
"loss": 0.788,
|
| 771 |
+
"step": 127
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 12.8,
|
| 775 |
+
"learning_rate": 6.061242437507131e-06,
|
| 776 |
+
"loss": 0.7914,
|
| 777 |
+
"step": 128
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"epoch": 12.9,
|
| 781 |
+
"learning_rate": 5.912918064264441e-06,
|
| 782 |
+
"loss": 0.8299,
|
| 783 |
+
"step": 129
|
| 784 |
+
},
|
| 785 |
+
{
|
| 786 |
+
"epoch": 13.0,
|
| 787 |
+
"learning_rate": 5.765665457425102e-06,
|
| 788 |
+
"loss": 0.7849,
|
| 789 |
+
"step": 130
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 13.1,
|
| 793 |
+
"learning_rate": 5.619523231433177e-06,
|
| 794 |
+
"loss": 0.7738,
|
| 795 |
+
"step": 131
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 13.2,
|
| 799 |
+
"learning_rate": 5.4745297095546125e-06,
|
| 800 |
+
"loss": 0.8217,
|
| 801 |
+
"step": 132
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 13.3,
|
| 805 |
+
"learning_rate": 5.330722913827594e-06,
|
| 806 |
+
"loss": 0.8051,
|
| 807 |
+
"step": 133
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 13.4,
|
| 811 |
+
"learning_rate": 5.18814055509195e-06,
|
| 812 |
+
"loss": 0.8071,
|
| 813 |
+
"step": 134
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 13.5,
|
| 817 |
+
"learning_rate": 5.046820023100129e-06,
|
| 818 |
+
"loss": 0.7114,
|
| 819 |
+
"step": 135
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 13.6,
|
| 823 |
+
"learning_rate": 4.9067983767123736e-06,
|
| 824 |
+
"loss": 0.7534,
|
| 825 |
+
"step": 136
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 13.7,
|
| 829 |
+
"learning_rate": 4.7681123341787e-06,
|
| 830 |
+
"loss": 0.7848,
|
| 831 |
+
"step": 137
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 13.8,
|
| 835 |
+
"learning_rate": 4.630798263510162e-06,
|
| 836 |
+
"loss": 0.7739,
|
| 837 |
+
"step": 138
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 13.9,
|
| 841 |
+
"learning_rate": 4.494892172941965e-06,
|
| 842 |
+
"loss": 0.7973,
|
| 843 |
+
"step": 139
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 14.0,
|
| 847 |
+
"learning_rate": 4.360429701490935e-06,
|
| 848 |
+
"loss": 0.8095,
|
| 849 |
+
"step": 140
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 14.1,
|
| 853 |
+
"learning_rate": 4.2274461096098085e-06,
|
| 854 |
+
"loss": 0.815,
|
| 855 |
+
"step": 141
|
| 856 |
+
},
|
| 857 |
+
{
|
| 858 |
+
"epoch": 14.2,
|
| 859 |
+
"learning_rate": 4.095976269940777e-06,
|
| 860 |
+
"loss": 0.7943,
|
| 861 |
+
"step": 142
|
| 862 |
+
},
|
| 863 |
+
{
|
| 864 |
+
"epoch": 14.3,
|
| 865 |
+
"learning_rate": 3.966054658170754e-06,
|
| 866 |
+
"loss": 0.7564,
|
| 867 |
+
"step": 143
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 14.4,
|
| 871 |
+
"learning_rate": 3.837715343990727e-06,
|
| 872 |
+
"loss": 0.803,
|
| 873 |
+
"step": 144
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 14.5,
|
| 877 |
+
"learning_rate": 3.7109919821615546e-06,
|
| 878 |
+
"loss": 0.7495,
|
| 879 |
+
"step": 145
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 14.6,
|
| 883 |
+
"learning_rate": 3.585917803688603e-06,
|
| 884 |
+
"loss": 0.7679,
|
| 885 |
+
"step": 146
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 14.7,
|
| 889 |
+
"learning_rate": 3.4625256071074776e-06,
|
| 890 |
+
"loss": 0.7697,
|
| 891 |
+
"step": 147
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 14.8,
|
| 895 |
+
"learning_rate": 3.3408477498831917e-06,
|
| 896 |
+
"loss": 0.7224,
|
| 897 |
+
"step": 148
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 14.9,
|
| 901 |
+
"learning_rate": 3.2209161399249677e-06,
|
| 902 |
+
"loss": 0.8048,
|
| 903 |
+
"step": 149
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 15.0,
|
| 907 |
+
"learning_rate": 3.1027622272189572e-06,
|
| 908 |
+
"loss": 0.7417,
|
| 909 |
+
"step": 150
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 15.1,
|
| 913 |
+
"learning_rate": 2.9864169955810085e-06,
|
| 914 |
+
"loss": 0.7383,
|
| 915 |
+
"step": 151
|
| 916 |
+
},
|
| 917 |
+
{
|
| 918 |
+
"epoch": 15.2,
|
| 919 |
+
"learning_rate": 2.8719109545317102e-06,
|
| 920 |
+
"loss": 0.7595,
|
| 921 |
+
"step": 152
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 15.3,
|
| 925 |
+
"learning_rate": 2.759274131295787e-06,
|
| 926 |
+
"loss": 0.7626,
|
| 927 |
+
"step": 153
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 15.4,
|
| 931 |
+
"learning_rate": 2.648536062927999e-06,
|
| 932 |
+
"loss": 0.7536,
|
| 933 |
+
"step": 154
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 15.5,
|
| 937 |
+
"learning_rate": 2.5397257885675396e-06,
|
| 938 |
+
"loss": 0.8178,
|
| 939 |
+
"step": 155
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 15.6,
|
| 943 |
+
"learning_rate": 2.432871841823047e-06,
|
| 944 |
+
"loss": 0.7955,
|
| 945 |
+
"step": 156
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 15.7,
|
| 949 |
+
"learning_rate": 2.328002243290138e-06,
|
| 950 |
+
"loss": 0.7337,
|
| 951 |
+
"step": 157
|
| 952 |
+
},
|
| 953 |
+
{
|
| 954 |
+
"epoch": 15.8,
|
| 955 |
+
"learning_rate": 2.2251444932035094e-06,
|
| 956 |
+
"loss": 0.7584,
|
| 957 |
+
"step": 158
|
| 958 |
+
},
|
| 959 |
+
{
|
| 960 |
+
"epoch": 15.9,
|
| 961 |
+
"learning_rate": 2.124325564225458e-06,
|
| 962 |
+
"loss": 0.7262,
|
| 963 |
+
"step": 159
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 16.0,
|
| 967 |
+
"learning_rate": 2.025571894372794e-06,
|
| 968 |
+
"loss": 0.8152,
|
| 969 |
+
"step": 160
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 16.1,
|
| 973 |
+
"learning_rate": 1.9289093800839067e-06,
|
| 974 |
+
"loss": 0.7813,
|
| 975 |
+
"step": 161
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 16.2,
|
| 979 |
+
"learning_rate": 1.8343633694278895e-06,
|
| 980 |
+
"loss": 0.7323,
|
| 981 |
+
"step": 162
|
| 982 |
+
},
|
| 983 |
+
{
|
| 984 |
+
"epoch": 16.3,
|
| 985 |
+
"learning_rate": 1.7419586554574364e-06,
|
| 986 |
+
"loss": 0.7492,
|
| 987 |
+
"step": 163
|
| 988 |
+
},
|
| 989 |
+
{
|
| 990 |
+
"epoch": 16.4,
|
| 991 |
+
"learning_rate": 1.6517194697072903e-06,
|
| 992 |
+
"loss": 0.7922,
|
| 993 |
+
"step": 164
|
| 994 |
+
},
|
| 995 |
+
{
|
| 996 |
+
"epoch": 16.5,
|
| 997 |
+
"learning_rate": 1.5636694758399563e-06,
|
| 998 |
+
"loss": 0.7663,
|
| 999 |
+
"step": 165
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 16.6,
|
| 1003 |
+
"learning_rate": 1.4778317634403082e-06,
|
| 1004 |
+
"loss": 0.7732,
|
| 1005 |
+
"step": 166
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 16.7,
|
| 1009 |
+
"learning_rate": 1.3942288419607476e-06,
|
| 1010 |
+
"loss": 0.8102,
|
| 1011 |
+
"step": 167
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 16.8,
|
| 1015 |
+
"learning_rate": 1.3128826348184886e-06,
|
| 1016 |
+
"loss": 0.7213,
|
| 1017 |
+
"step": 168
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 16.9,
|
| 1021 |
+
"learning_rate": 1.233814473646524e-06,
|
| 1022 |
+
"loss": 0.7614,
|
| 1023 |
+
"step": 169
|
| 1024 |
+
},
|
| 1025 |
+
{
|
| 1026 |
+
"epoch": 17.0,
|
| 1027 |
+
"learning_rate": 1.1570450926997657e-06,
|
| 1028 |
+
"loss": 0.7527,
|
| 1029 |
+
"step": 170
|
| 1030 |
+
},
|
| 1031 |
+
{
|
| 1032 |
+
"epoch": 17.1,
|
| 1033 |
+
"learning_rate": 1.0825946234178575e-06,
|
| 1034 |
+
"loss": 0.7594,
|
| 1035 |
+
"step": 171
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 17.2,
|
| 1039 |
+
"learning_rate": 1.010482589146048e-06,
|
| 1040 |
+
"loss": 0.7954,
|
| 1041 |
+
"step": 172
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 17.3,
|
| 1045 |
+
"learning_rate": 9.407279000155311e-07,
|
| 1046 |
+
"loss": 0.7166,
|
| 1047 |
+
"step": 173
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 17.4,
|
| 1051 |
+
"learning_rate": 8.733488479845997e-07,
|
| 1052 |
+
"loss": 0.7717,
|
| 1053 |
+
"step": 174
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 17.5,
|
| 1057 |
+
"learning_rate": 8.083631020418792e-07,
|
| 1058 |
+
"loss": 0.7513,
|
| 1059 |
+
"step": 175
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 17.6,
|
| 1063 |
+
"learning_rate": 7.457877035729588e-07,
|
| 1064 |
+
"loss": 0.7614,
|
| 1065 |
+
"step": 176
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 17.7,
|
| 1069 |
+
"learning_rate": 6.856390618915775e-07,
|
| 1070 |
+
"loss": 0.7488,
|
| 1071 |
+
"step": 177
|
| 1072 |
+
},
|
| 1073 |
+
{
|
| 1074 |
+
"epoch": 17.8,
|
| 1075 |
+
"learning_rate": 6.279329499365649e-07,
|
| 1076 |
+
"loss": 0.7793,
|
| 1077 |
+
"step": 178
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 17.9,
|
| 1081 |
+
"learning_rate": 5.726845001356573e-07,
|
| 1082 |
+
"loss": 0.7329,
|
| 1083 |
+
"step": 179
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 18.0,
|
| 1087 |
+
"learning_rate": 5.199082004372958e-07,
|
| 1088 |
+
"loss": 0.7695,
|
| 1089 |
+
"step": 180
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 18.1,
|
| 1093 |
+
"learning_rate": 4.696178905113913e-07,
|
| 1094 |
+
"loss": 0.7533,
|
| 1095 |
+
"step": 181
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 18.2,
|
| 1099 |
+
"learning_rate": 4.218267581201296e-07,
|
| 1100 |
+
"loss": 0.7467,
|
| 1101 |
+
"step": 182
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 18.3,
|
| 1105 |
+
"learning_rate": 3.7654733565969826e-07,
|
| 1106 |
+
"loss": 0.7375,
|
| 1107 |
+
"step": 183
|
| 1108 |
+
},
|
| 1109 |
+
{
|
| 1110 |
+
"epoch": 18.4,
|
| 1111 |
+
"learning_rate": 3.3379149687388866e-07,
|
| 1112 |
+
"loss": 0.7823,
|
| 1113 |
+
"step": 184
|
| 1114 |
+
},
|
| 1115 |
+
{
|
| 1116 |
+
"epoch": 18.5,
|
| 1117 |
+
"learning_rate": 2.935704537404083e-07,
|
| 1118 |
+
"loss": 0.7461,
|
| 1119 |
+
"step": 185
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 18.6,
|
| 1123 |
+
"learning_rate": 2.5589475353073987e-07,
|
| 1124 |
+
"loss": 0.7944,
|
| 1125 |
+
"step": 186
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 18.7,
|
| 1129 |
+
"learning_rate": 2.2077427604429435e-07,
|
| 1130 |
+
"loss": 0.7583,
|
| 1131 |
+
"step": 187
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 18.8,
|
| 1135 |
+
"learning_rate": 1.8821823101760949e-07,
|
| 1136 |
+
"loss": 0.792,
|
| 1137 |
+
"step": 188
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 18.9,
|
| 1141 |
+
"learning_rate": 1.5823515570925763e-07,
|
| 1142 |
+
"loss": 0.7359,
|
| 1143 |
+
"step": 189
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 19.0,
|
| 1147 |
+
"learning_rate": 1.30832912661093e-07,
|
| 1148 |
+
"loss": 0.749,
|
| 1149 |
+
"step": 190
|
| 1150 |
+
},
|
| 1151 |
+
{
|
| 1152 |
+
"epoch": 19.1,
|
| 1153 |
+
"learning_rate": 1.0601868763643997e-07,
|
| 1154 |
+
"loss": 0.7866,
|
| 1155 |
+
"step": 191
|
| 1156 |
+
},
|
| 1157 |
+
{
|
| 1158 |
+
"epoch": 19.2,
|
| 1159 |
+
"learning_rate": 8.379898773574924e-08,
|
| 1160 |
+
"loss": 0.711,
|
| 1161 |
+
"step": 192
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 19.3,
|
| 1165 |
+
"learning_rate": 6.417963969022389e-08,
|
| 1166 |
+
"loss": 0.7756,
|
| 1167 |
+
"step": 193
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 19.4,
|
| 1171 |
+
"learning_rate": 4.716578833386054e-08,
|
| 1172 |
+
"loss": 0.7339,
|
| 1173 |
+
"step": 194
|
| 1174 |
+
},
|
| 1175 |
+
{
|
| 1176 |
+
"epoch": 19.5,
|
| 1177 |
+
"learning_rate": 3.2761895254306285e-08,
|
| 1178 |
+
"loss": 0.7586,
|
| 1179 |
+
"step": 195
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 19.6,
|
| 1183 |
+
"learning_rate": 2.0971737622883515e-08,
|
| 1184 |
+
"loss": 0.7735,
|
| 1185 |
+
"step": 196
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 19.7,
|
| 1189 |
+
"learning_rate": 1.179840720409331e-08,
|
| 1190 |
+
"loss": 0.7166,
|
| 1191 |
+
"step": 197
|
| 1192 |
+
},
|
| 1193 |
+
{
|
| 1194 |
+
"epoch": 19.8,
|
| 1195 |
+
"learning_rate": 5.2443095448506674e-09,
|
| 1196 |
+
"loss": 0.7639,
|
| 1197 |
+
"step": 198
|
| 1198 |
+
},
|
| 1199 |
+
{
|
| 1200 |
+
"epoch": 19.9,
|
| 1201 |
+
"learning_rate": 1.3111633436779792e-09,
|
| 1202 |
+
"loss": 0.7625,
|
| 1203 |
+
"step": 199
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 20.0,
|
| 1207 |
+
"learning_rate": 0.0,
|
| 1208 |
+
"loss": 0.7749,
|
| 1209 |
+
"step": 200
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 20.0,
|
| 1213 |
+
"step": 200,
|
| 1214 |
+
"total_flos": 16748476579840.0,
|
| 1215 |
+
"train_loss": 0.9612581425905228,
|
| 1216 |
+
"train_runtime": 1726.3908,
|
| 1217 |
+
"train_samples_per_second": 3.707,
|
| 1218 |
+
"train_steps_per_second": 0.116
|
| 1219 |
+
}
|
| 1220 |
+
],
|
| 1221 |
+
"logging_steps": 1.0,
|
| 1222 |
+
"max_steps": 200,
|
| 1223 |
+
"num_input_tokens_seen": 0,
|
| 1224 |
+
"num_train_epochs": 20,
|
| 1225 |
+
"save_steps": 50000,
|
| 1226 |
+
"total_flos": 16748476579840.0,
|
| 1227 |
+
"train_batch_size": 8,
|
| 1228 |
+
"trial_name": null,
|
| 1229 |
+
"trial_params": null
|
| 1230 |
+
}
|