Add files using upload-large-folder tool
Browse files- added_tokens.json +37 -0
- chat_template.jinja +6 -0
- config.json +123 -0
- merges.txt +0 -0
- model.safetensors +3 -0
- special_tokens_map.json +40 -0
- tokenizer_config.json +322 -0
- trainer_state.json +1834 -0
- training_args.bin +3 -0
- vocab.json +0 -0
added_tokens.json
ADDED
|
@@ -0,0 +1,37 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"</box>": 151677,
|
| 3 |
+
"</img>": 151670,
|
| 4 |
+
"</quad>": 151673,
|
| 5 |
+
"</ref>": 151675,
|
| 6 |
+
"</think>": 151668,
|
| 7 |
+
"</tool_call>": 151658,
|
| 8 |
+
"</tool_response>": 151666,
|
| 9 |
+
"<IMG_CONTEXT>": 151671,
|
| 10 |
+
"<box>": 151676,
|
| 11 |
+
"<img>": 151669,
|
| 12 |
+
"<quad>": 151672,
|
| 13 |
+
"<ref>": 151674,
|
| 14 |
+
"<think>": 151667,
|
| 15 |
+
"<tool_call>": 151657,
|
| 16 |
+
"<tool_response>": 151665,
|
| 17 |
+
"<|box_end|>": 151649,
|
| 18 |
+
"<|box_start|>": 151648,
|
| 19 |
+
"<|endoftext|>": 151643,
|
| 20 |
+
"<|file_sep|>": 151664,
|
| 21 |
+
"<|fim_middle|>": 151660,
|
| 22 |
+
"<|fim_pad|>": 151662,
|
| 23 |
+
"<|fim_prefix|>": 151659,
|
| 24 |
+
"<|fim_suffix|>": 151661,
|
| 25 |
+
"<|im_end|>": 151645,
|
| 26 |
+
"<|im_start|>": 151644,
|
| 27 |
+
"<|image_pad|>": 151655,
|
| 28 |
+
"<|object_ref_end|>": 151647,
|
| 29 |
+
"<|object_ref_start|>": 151646,
|
| 30 |
+
"<|quad_end|>": 151651,
|
| 31 |
+
"<|quad_start|>": 151650,
|
| 32 |
+
"<|repo_name|>": 151663,
|
| 33 |
+
"<|video_pad|>": 151656,
|
| 34 |
+
"<|vision_end|>": 151653,
|
| 35 |
+
"<|vision_pad|>": 151654,
|
| 36 |
+
"<|vision_start|>": 151652
|
| 37 |
+
}
|
chat_template.jinja
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{% for message in messages %}{{'<|im_start|>' + message['role'] + '
|
| 2 |
+
'}}{% if message['content'] is string %}{{ message['content'] }}{% else %}{% for content in message['content'] %}{% if content['type'] == 'image' %}{{ '<image>
|
| 3 |
+
' }}{% elif content['type'] == 'video' %}{{ '<video>
|
| 4 |
+
' }}{% elif content['type'] == 'text' %}{{ content['text'] }}{% endif %}{% endfor %}{% endif %}{{'<|im_end|>
|
| 5 |
+
'}}{% endfor %}{% if add_generation_prompt %}{{'<|im_start|>assistant
|
| 6 |
+
' }}{% endif %}
|
config.json
ADDED
|
@@ -0,0 +1,123 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"InternVLChatModel"
|
| 4 |
+
],
|
| 5 |
+
"auto_map": {
|
| 6 |
+
"AutoConfig": "configuration_internvl_chat.InternVLChatConfig",
|
| 7 |
+
"AutoModel": "modeling_internvl_chat.InternVLChatModel",
|
| 8 |
+
"AutoModelForCausalLM": "modeling_internvl_chat.InternVLChatModel"
|
| 9 |
+
},
|
| 10 |
+
"downsample_ratio": 0.5,
|
| 11 |
+
"dtype": "bfloat16",
|
| 12 |
+
"dynamic_image_size": true,
|
| 13 |
+
"eos_token_id": 151645,
|
| 14 |
+
"force_image_size": 336,
|
| 15 |
+
"llm_config": {
|
| 16 |
+
"_name_or_path": "/root/codespace/checkpoints/Qwen3-1.7B",
|
| 17 |
+
"architectures": [
|
| 18 |
+
"Qwen3ForCausalLM"
|
| 19 |
+
],
|
| 20 |
+
"attention_bias": false,
|
| 21 |
+
"attention_dropout": 0.0,
|
| 22 |
+
"bos_token_id": 151643,
|
| 23 |
+
"debug": false,
|
| 24 |
+
"dtype": "bfloat16",
|
| 25 |
+
"eos_token_id": 151645,
|
| 26 |
+
"ep_size": 1,
|
| 27 |
+
"head_dim": 128,
|
| 28 |
+
"hidden_act": "silu",
|
| 29 |
+
"hidden_size": 2048,
|
| 30 |
+
"initializer_range": 0.02,
|
| 31 |
+
"intermediate_size": 6144,
|
| 32 |
+
"layer_types": [
|
| 33 |
+
"full_attention",
|
| 34 |
+
"full_attention",
|
| 35 |
+
"full_attention",
|
| 36 |
+
"full_attention",
|
| 37 |
+
"full_attention",
|
| 38 |
+
"full_attention",
|
| 39 |
+
"full_attention",
|
| 40 |
+
"full_attention",
|
| 41 |
+
"full_attention",
|
| 42 |
+
"full_attention",
|
| 43 |
+
"full_attention",
|
| 44 |
+
"full_attention",
|
| 45 |
+
"full_attention",
|
| 46 |
+
"full_attention",
|
| 47 |
+
"full_attention",
|
| 48 |
+
"full_attention",
|
| 49 |
+
"full_attention",
|
| 50 |
+
"full_attention",
|
| 51 |
+
"full_attention",
|
| 52 |
+
"full_attention",
|
| 53 |
+
"full_attention",
|
| 54 |
+
"full_attention",
|
| 55 |
+
"full_attention",
|
| 56 |
+
"full_attention",
|
| 57 |
+
"full_attention",
|
| 58 |
+
"full_attention",
|
| 59 |
+
"full_attention",
|
| 60 |
+
"full_attention"
|
| 61 |
+
],
|
| 62 |
+
"max_position_embeddings": 40960,
|
| 63 |
+
"max_window_layers": 28,
|
| 64 |
+
"micro_forward": false,
|
| 65 |
+
"model_type": "qwen3",
|
| 66 |
+
"num_attention_heads": 16,
|
| 67 |
+
"num_hidden_layers": 28,
|
| 68 |
+
"num_key_value_heads": 8,
|
| 69 |
+
"rms_norm_eps": 1e-06,
|
| 70 |
+
"rope_scaling": null,
|
| 71 |
+
"rope_theta": 1000000,
|
| 72 |
+
"skip_checkpoint": false,
|
| 73 |
+
"sliding_window": null,
|
| 74 |
+
"use_cache": false,
|
| 75 |
+
"use_deepep": false,
|
| 76 |
+
"use_sliding_window": false,
|
| 77 |
+
"vocab_size": 151936
|
| 78 |
+
},
|
| 79 |
+
"max_dynamic_patch": 12,
|
| 80 |
+
"min_dynamic_patch": 1,
|
| 81 |
+
"model_type": "internvl_chat",
|
| 82 |
+
"output_attentions": false,
|
| 83 |
+
"pad2square": false,
|
| 84 |
+
"pad_token_id": 151643,
|
| 85 |
+
"ps_version": "v2",
|
| 86 |
+
"select_layer": -1,
|
| 87 |
+
"template": "internvl2_5",
|
| 88 |
+
"tie_word_embeddings": false,
|
| 89 |
+
"transformers_version": null,
|
| 90 |
+
"use_backbone_lora": 0,
|
| 91 |
+
"use_llm_lora": 0,
|
| 92 |
+
"use_thumbnail": true,
|
| 93 |
+
"vision_config": {
|
| 94 |
+
"architectures": [
|
| 95 |
+
"InternVisionModel"
|
| 96 |
+
],
|
| 97 |
+
"attention_dropout": 0.0,
|
| 98 |
+
"auto_map": {
|
| 99 |
+
"AutoConfig": "configuration_intern_vit.InternVisionConfig",
|
| 100 |
+
"AutoModel": "modeling_intern_vit.InternVisionModel"
|
| 101 |
+
},
|
| 102 |
+
"drop_path_rate": 0.0,
|
| 103 |
+
"dropout": 0.0,
|
| 104 |
+
"dtype": "bfloat16",
|
| 105 |
+
"hidden_act": "gelu",
|
| 106 |
+
"hidden_size": 1024,
|
| 107 |
+
"image_size": 448,
|
| 108 |
+
"initializer_factor": 1.0,
|
| 109 |
+
"initializer_range": 0.02,
|
| 110 |
+
"intermediate_size": 4096,
|
| 111 |
+
"layer_norm_eps": 1e-06,
|
| 112 |
+
"model_type": "intern_vit_6b",
|
| 113 |
+
"norm_type": "layer_norm",
|
| 114 |
+
"num_attention_heads": 16,
|
| 115 |
+
"num_channels": 3,
|
| 116 |
+
"num_hidden_layers": 24,
|
| 117 |
+
"patch_size": 14,
|
| 118 |
+
"qk_normalization": false,
|
| 119 |
+
"qkv_bias": true,
|
| 120 |
+
"use_fa3": false,
|
| 121 |
+
"use_flash_attn": false
|
| 122 |
+
}
|
| 123 |
+
}
|
merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a10a7481992249adc325493e244edae46bafe717ed404e5a4d0b509aaef78372
|
| 3 |
+
size 4092735128
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,40 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"additional_special_tokens": [
|
| 3 |
+
"<|im_start|>",
|
| 4 |
+
"<|im_end|>",
|
| 5 |
+
"<|object_ref_start|>",
|
| 6 |
+
"<|object_ref_end|>",
|
| 7 |
+
"<|box_start|>",
|
| 8 |
+
"<|box_end|>",
|
| 9 |
+
"<|quad_start|>",
|
| 10 |
+
"<|quad_end|>",
|
| 11 |
+
"<|vision_start|>",
|
| 12 |
+
"<|vision_end|>",
|
| 13 |
+
"<|vision_pad|>",
|
| 14 |
+
"<|image_pad|>",
|
| 15 |
+
"<|video_pad|>",
|
| 16 |
+
"<img>",
|
| 17 |
+
"</img>",
|
| 18 |
+
"<IMG_CONTEXT>",
|
| 19 |
+
"<quad>",
|
| 20 |
+
"</quad>",
|
| 21 |
+
"<ref>",
|
| 22 |
+
"</ref>",
|
| 23 |
+
"<box>",
|
| 24 |
+
"</box>"
|
| 25 |
+
],
|
| 26 |
+
"eos_token": {
|
| 27 |
+
"content": "<|im_end|>",
|
| 28 |
+
"lstrip": false,
|
| 29 |
+
"normalized": false,
|
| 30 |
+
"rstrip": false,
|
| 31 |
+
"single_word": false
|
| 32 |
+
},
|
| 33 |
+
"pad_token": {
|
| 34 |
+
"content": "<|endoftext|>",
|
| 35 |
+
"lstrip": false,
|
| 36 |
+
"normalized": false,
|
| 37 |
+
"rstrip": false,
|
| 38 |
+
"single_word": false
|
| 39 |
+
}
|
| 40 |
+
}
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,322 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_bos_token": false,
|
| 3 |
+
"add_eos_token": false,
|
| 4 |
+
"add_prefix_space": false,
|
| 5 |
+
"added_tokens_decoder": {
|
| 6 |
+
"151643": {
|
| 7 |
+
"content": "<|endoftext|>",
|
| 8 |
+
"lstrip": false,
|
| 9 |
+
"normalized": false,
|
| 10 |
+
"rstrip": false,
|
| 11 |
+
"single_word": false,
|
| 12 |
+
"special": true
|
| 13 |
+
},
|
| 14 |
+
"151644": {
|
| 15 |
+
"content": "<|im_start|>",
|
| 16 |
+
"lstrip": false,
|
| 17 |
+
"normalized": false,
|
| 18 |
+
"rstrip": false,
|
| 19 |
+
"single_word": false,
|
| 20 |
+
"special": true
|
| 21 |
+
},
|
| 22 |
+
"151645": {
|
| 23 |
+
"content": "<|im_end|>",
|
| 24 |
+
"lstrip": false,
|
| 25 |
+
"normalized": false,
|
| 26 |
+
"rstrip": false,
|
| 27 |
+
"single_word": false,
|
| 28 |
+
"special": true
|
| 29 |
+
},
|
| 30 |
+
"151646": {
|
| 31 |
+
"content": "<|object_ref_start|>",
|
| 32 |
+
"lstrip": false,
|
| 33 |
+
"normalized": false,
|
| 34 |
+
"rstrip": false,
|
| 35 |
+
"single_word": false,
|
| 36 |
+
"special": true
|
| 37 |
+
},
|
| 38 |
+
"151647": {
|
| 39 |
+
"content": "<|object_ref_end|>",
|
| 40 |
+
"lstrip": false,
|
| 41 |
+
"normalized": false,
|
| 42 |
+
"rstrip": false,
|
| 43 |
+
"single_word": false,
|
| 44 |
+
"special": true
|
| 45 |
+
},
|
| 46 |
+
"151648": {
|
| 47 |
+
"content": "<|box_start|>",
|
| 48 |
+
"lstrip": false,
|
| 49 |
+
"normalized": false,
|
| 50 |
+
"rstrip": false,
|
| 51 |
+
"single_word": false,
|
| 52 |
+
"special": true
|
| 53 |
+
},
|
| 54 |
+
"151649": {
|
| 55 |
+
"content": "<|box_end|>",
|
| 56 |
+
"lstrip": false,
|
| 57 |
+
"normalized": false,
|
| 58 |
+
"rstrip": false,
|
| 59 |
+
"single_word": false,
|
| 60 |
+
"special": true
|
| 61 |
+
},
|
| 62 |
+
"151650": {
|
| 63 |
+
"content": "<|quad_start|>",
|
| 64 |
+
"lstrip": false,
|
| 65 |
+
"normalized": false,
|
| 66 |
+
"rstrip": false,
|
| 67 |
+
"single_word": false,
|
| 68 |
+
"special": true
|
| 69 |
+
},
|
| 70 |
+
"151651": {
|
| 71 |
+
"content": "<|quad_end|>",
|
| 72 |
+
"lstrip": false,
|
| 73 |
+
"normalized": false,
|
| 74 |
+
"rstrip": false,
|
| 75 |
+
"single_word": false,
|
| 76 |
+
"special": true
|
| 77 |
+
},
|
| 78 |
+
"151652": {
|
| 79 |
+
"content": "<|vision_start|>",
|
| 80 |
+
"lstrip": false,
|
| 81 |
+
"normalized": false,
|
| 82 |
+
"rstrip": false,
|
| 83 |
+
"single_word": false,
|
| 84 |
+
"special": true
|
| 85 |
+
},
|
| 86 |
+
"151653": {
|
| 87 |
+
"content": "<|vision_end|>",
|
| 88 |
+
"lstrip": false,
|
| 89 |
+
"normalized": false,
|
| 90 |
+
"rstrip": false,
|
| 91 |
+
"single_word": false,
|
| 92 |
+
"special": true
|
| 93 |
+
},
|
| 94 |
+
"151654": {
|
| 95 |
+
"content": "<|vision_pad|>",
|
| 96 |
+
"lstrip": false,
|
| 97 |
+
"normalized": false,
|
| 98 |
+
"rstrip": false,
|
| 99 |
+
"single_word": false,
|
| 100 |
+
"special": true
|
| 101 |
+
},
|
| 102 |
+
"151655": {
|
| 103 |
+
"content": "<|image_pad|>",
|
| 104 |
+
"lstrip": false,
|
| 105 |
+
"normalized": false,
|
| 106 |
+
"rstrip": false,
|
| 107 |
+
"single_word": false,
|
| 108 |
+
"special": true
|
| 109 |
+
},
|
| 110 |
+
"151656": {
|
| 111 |
+
"content": "<|video_pad|>",
|
| 112 |
+
"lstrip": false,
|
| 113 |
+
"normalized": false,
|
| 114 |
+
"rstrip": false,
|
| 115 |
+
"single_word": false,
|
| 116 |
+
"special": true
|
| 117 |
+
},
|
| 118 |
+
"151657": {
|
| 119 |
+
"content": "<tool_call>",
|
| 120 |
+
"lstrip": false,
|
| 121 |
+
"normalized": false,
|
| 122 |
+
"rstrip": false,
|
| 123 |
+
"single_word": false,
|
| 124 |
+
"special": false
|
| 125 |
+
},
|
| 126 |
+
"151658": {
|
| 127 |
+
"content": "</tool_call>",
|
| 128 |
+
"lstrip": false,
|
| 129 |
+
"normalized": false,
|
| 130 |
+
"rstrip": false,
|
| 131 |
+
"single_word": false,
|
| 132 |
+
"special": false
|
| 133 |
+
},
|
| 134 |
+
"151659": {
|
| 135 |
+
"content": "<|fim_prefix|>",
|
| 136 |
+
"lstrip": false,
|
| 137 |
+
"normalized": false,
|
| 138 |
+
"rstrip": false,
|
| 139 |
+
"single_word": false,
|
| 140 |
+
"special": false
|
| 141 |
+
},
|
| 142 |
+
"151660": {
|
| 143 |
+
"content": "<|fim_middle|>",
|
| 144 |
+
"lstrip": false,
|
| 145 |
+
"normalized": false,
|
| 146 |
+
"rstrip": false,
|
| 147 |
+
"single_word": false,
|
| 148 |
+
"special": false
|
| 149 |
+
},
|
| 150 |
+
"151661": {
|
| 151 |
+
"content": "<|fim_suffix|>",
|
| 152 |
+
"lstrip": false,
|
| 153 |
+
"normalized": false,
|
| 154 |
+
"rstrip": false,
|
| 155 |
+
"single_word": false,
|
| 156 |
+
"special": false
|
| 157 |
+
},
|
| 158 |
+
"151662": {
|
| 159 |
+
"content": "<|fim_pad|>",
|
| 160 |
+
"lstrip": false,
|
| 161 |
+
"normalized": false,
|
| 162 |
+
"rstrip": false,
|
| 163 |
+
"single_word": false,
|
| 164 |
+
"special": false
|
| 165 |
+
},
|
| 166 |
+
"151663": {
|
| 167 |
+
"content": "<|repo_name|>",
|
| 168 |
+
"lstrip": false,
|
| 169 |
+
"normalized": false,
|
| 170 |
+
"rstrip": false,
|
| 171 |
+
"single_word": false,
|
| 172 |
+
"special": false
|
| 173 |
+
},
|
| 174 |
+
"151664": {
|
| 175 |
+
"content": "<|file_sep|>",
|
| 176 |
+
"lstrip": false,
|
| 177 |
+
"normalized": false,
|
| 178 |
+
"rstrip": false,
|
| 179 |
+
"single_word": false,
|
| 180 |
+
"special": false
|
| 181 |
+
},
|
| 182 |
+
"151665": {
|
| 183 |
+
"content": "<tool_response>",
|
| 184 |
+
"lstrip": false,
|
| 185 |
+
"normalized": false,
|
| 186 |
+
"rstrip": false,
|
| 187 |
+
"single_word": false,
|
| 188 |
+
"special": false
|
| 189 |
+
},
|
| 190 |
+
"151666": {
|
| 191 |
+
"content": "</tool_response>",
|
| 192 |
+
"lstrip": false,
|
| 193 |
+
"normalized": false,
|
| 194 |
+
"rstrip": false,
|
| 195 |
+
"single_word": false,
|
| 196 |
+
"special": false
|
| 197 |
+
},
|
| 198 |
+
"151667": {
|
| 199 |
+
"content": "<think>",
|
| 200 |
+
"lstrip": false,
|
| 201 |
+
"normalized": false,
|
| 202 |
+
"rstrip": false,
|
| 203 |
+
"single_word": false,
|
| 204 |
+
"special": false
|
| 205 |
+
},
|
| 206 |
+
"151668": {
|
| 207 |
+
"content": "</think>",
|
| 208 |
+
"lstrip": false,
|
| 209 |
+
"normalized": false,
|
| 210 |
+
"rstrip": false,
|
| 211 |
+
"single_word": false,
|
| 212 |
+
"special": false
|
| 213 |
+
},
|
| 214 |
+
"151669": {
|
| 215 |
+
"content": "<img>",
|
| 216 |
+
"lstrip": false,
|
| 217 |
+
"normalized": false,
|
| 218 |
+
"rstrip": false,
|
| 219 |
+
"single_word": false,
|
| 220 |
+
"special": true
|
| 221 |
+
},
|
| 222 |
+
"151670": {
|
| 223 |
+
"content": "</img>",
|
| 224 |
+
"lstrip": false,
|
| 225 |
+
"normalized": false,
|
| 226 |
+
"rstrip": false,
|
| 227 |
+
"single_word": false,
|
| 228 |
+
"special": true
|
| 229 |
+
},
|
| 230 |
+
"151671": {
|
| 231 |
+
"content": "<IMG_CONTEXT>",
|
| 232 |
+
"lstrip": false,
|
| 233 |
+
"normalized": false,
|
| 234 |
+
"rstrip": false,
|
| 235 |
+
"single_word": false,
|
| 236 |
+
"special": true
|
| 237 |
+
},
|
| 238 |
+
"151672": {
|
| 239 |
+
"content": "<quad>",
|
| 240 |
+
"lstrip": false,
|
| 241 |
+
"normalized": false,
|
| 242 |
+
"rstrip": false,
|
| 243 |
+
"single_word": false,
|
| 244 |
+
"special": true
|
| 245 |
+
},
|
| 246 |
+
"151673": {
|
| 247 |
+
"content": "</quad>",
|
| 248 |
+
"lstrip": false,
|
| 249 |
+
"normalized": false,
|
| 250 |
+
"rstrip": false,
|
| 251 |
+
"single_word": false,
|
| 252 |
+
"special": true
|
| 253 |
+
},
|
| 254 |
+
"151674": {
|
| 255 |
+
"content": "<ref>",
|
| 256 |
+
"lstrip": false,
|
| 257 |
+
"normalized": false,
|
| 258 |
+
"rstrip": false,
|
| 259 |
+
"single_word": false,
|
| 260 |
+
"special": true
|
| 261 |
+
},
|
| 262 |
+
"151675": {
|
| 263 |
+
"content": "</ref>",
|
| 264 |
+
"lstrip": false,
|
| 265 |
+
"normalized": false,
|
| 266 |
+
"rstrip": false,
|
| 267 |
+
"single_word": false,
|
| 268 |
+
"special": true
|
| 269 |
+
},
|
| 270 |
+
"151676": {
|
| 271 |
+
"content": "<box>",
|
| 272 |
+
"lstrip": false,
|
| 273 |
+
"normalized": false,
|
| 274 |
+
"rstrip": false,
|
| 275 |
+
"single_word": false,
|
| 276 |
+
"special": true
|
| 277 |
+
},
|
| 278 |
+
"151677": {
|
| 279 |
+
"content": "</box>",
|
| 280 |
+
"lstrip": false,
|
| 281 |
+
"normalized": false,
|
| 282 |
+
"rstrip": false,
|
| 283 |
+
"single_word": false,
|
| 284 |
+
"special": true
|
| 285 |
+
}
|
| 286 |
+
},
|
| 287 |
+
"additional_special_tokens": [
|
| 288 |
+
"<|im_start|>",
|
| 289 |
+
"<|im_end|>",
|
| 290 |
+
"<|object_ref_start|>",
|
| 291 |
+
"<|object_ref_end|>",
|
| 292 |
+
"<|box_start|>",
|
| 293 |
+
"<|box_end|>",
|
| 294 |
+
"<|quad_start|>",
|
| 295 |
+
"<|quad_end|>",
|
| 296 |
+
"<|vision_start|>",
|
| 297 |
+
"<|vision_end|>",
|
| 298 |
+
"<|vision_pad|>",
|
| 299 |
+
"<|image_pad|>",
|
| 300 |
+
"<|video_pad|>",
|
| 301 |
+
"<img>",
|
| 302 |
+
"</img>",
|
| 303 |
+
"<IMG_CONTEXT>",
|
| 304 |
+
"<quad>",
|
| 305 |
+
"</quad>",
|
| 306 |
+
"<ref>",
|
| 307 |
+
"</ref>",
|
| 308 |
+
"<box>",
|
| 309 |
+
"</box>"
|
| 310 |
+
],
|
| 311 |
+
"bos_token": null,
|
| 312 |
+
"clean_up_tokenization_spaces": false,
|
| 313 |
+
"eos_token": "<|im_end|>",
|
| 314 |
+
"errors": "replace",
|
| 315 |
+
"extra_special_tokens": {},
|
| 316 |
+
"model_max_length": 14588,
|
| 317 |
+
"pad_token": "<|endoftext|>",
|
| 318 |
+
"processor_class": "InternVLProcessor",
|
| 319 |
+
"split_special_tokens": false,
|
| 320 |
+
"tokenizer_class": "Qwen2Tokenizer",
|
| 321 |
+
"unk_token": null
|
| 322 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1834 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.6944444444444444,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 200,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"avg_label": 0.45932112634181976,
|
| 14 |
+
"epoch": 0.003472222222222222,
|
| 15 |
+
"grad_norm": 0.515625,
|
| 16 |
+
"learning_rate": 0.0,
|
| 17 |
+
"loss": 0.02412083954550326,
|
| 18 |
+
"num_pairs": 28.0,
|
| 19 |
+
"step": 1
|
| 20 |
+
},
|
| 21 |
+
{
|
| 22 |
+
"avg_label": 0.4166666865348816,
|
| 23 |
+
"epoch": 0.006944444444444444,
|
| 24 |
+
"grad_norm": 11.625,
|
| 25 |
+
"learning_rate": 6.896551724137931e-07,
|
| 26 |
+
"loss": 0.40197963267564774,
|
| 27 |
+
"num_pairs": 21.5,
|
| 28 |
+
"step": 2
|
| 29 |
+
},
|
| 30 |
+
{
|
| 31 |
+
"avg_label": 0.4724903553724289,
|
| 32 |
+
"epoch": 0.010416666666666666,
|
| 33 |
+
"grad_norm": 13.25,
|
| 34 |
+
"learning_rate": 1.3793103448275862e-06,
|
| 35 |
+
"loss": 0.5004251897335052,
|
| 36 |
+
"num_pairs": 28.0,
|
| 37 |
+
"step": 3
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"avg_label": 0.44384056329727173,
|
| 41 |
+
"epoch": 0.013888888888888888,
|
| 42 |
+
"grad_norm": 5.875,
|
| 43 |
+
"learning_rate": 2.0689655172413796e-06,
|
| 44 |
+
"loss": 0.23160110414028168,
|
| 45 |
+
"num_pairs": 27.5,
|
| 46 |
+
"step": 4
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"avg_label": 0.4630681872367859,
|
| 50 |
+
"epoch": 0.017361111111111112,
|
| 51 |
+
"grad_norm": 8.9375,
|
| 52 |
+
"learning_rate": 2.7586206896551725e-06,
|
| 53 |
+
"loss": 0.4468539021909237,
|
| 54 |
+
"num_pairs": 24.0,
|
| 55 |
+
"step": 5
|
| 56 |
+
},
|
| 57 |
+
{
|
| 58 |
+
"avg_label": 0.45495015382766724,
|
| 59 |
+
"epoch": 0.020833333333333332,
|
| 60 |
+
"grad_norm": 1.34375,
|
| 61 |
+
"learning_rate": 3.448275862068966e-06,
|
| 62 |
+
"loss": 0.017617323901504278,
|
| 63 |
+
"num_pairs": 28.0,
|
| 64 |
+
"step": 6
|
| 65 |
+
},
|
| 66 |
+
{
|
| 67 |
+
"avg_label": 0.4732142984867096,
|
| 68 |
+
"epoch": 0.024305555555555556,
|
| 69 |
+
"grad_norm": 11.5625,
|
| 70 |
+
"learning_rate": 4.137931034482759e-06,
|
| 71 |
+
"loss": 0.4206290766596794,
|
| 72 |
+
"num_pairs": 24.0,
|
| 73 |
+
"step": 7
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"avg_label": 0.4736842066049576,
|
| 77 |
+
"epoch": 0.027777777777777776,
|
| 78 |
+
"grad_norm": 0.1484375,
|
| 79 |
+
"learning_rate": 4.8275862068965525e-06,
|
| 80 |
+
"loss": 0.004913174081593752,
|
| 81 |
+
"num_pairs": 27.0,
|
| 82 |
+
"step": 8
|
| 83 |
+
},
|
| 84 |
+
{
|
| 85 |
+
"avg_label": 0.45513392984867096,
|
| 86 |
+
"epoch": 0.03125,
|
| 87 |
+
"grad_norm": 7.84375,
|
| 88 |
+
"learning_rate": 5.517241379310345e-06,
|
| 89 |
+
"loss": 0.40890943817794323,
|
| 90 |
+
"num_pairs": 28.0,
|
| 91 |
+
"step": 9
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"avg_label": 0.4797794222831726,
|
| 95 |
+
"epoch": 0.034722222222222224,
|
| 96 |
+
"grad_norm": 4.21875,
|
| 97 |
+
"learning_rate": 6.206896551724138e-06,
|
| 98 |
+
"loss": 0.07930763380136341,
|
| 99 |
+
"num_pairs": 28.0,
|
| 100 |
+
"step": 10
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"avg_label": 0.4505208283662796,
|
| 104 |
+
"epoch": 0.03819444444444445,
|
| 105 |
+
"grad_norm": 18.0,
|
| 106 |
+
"learning_rate": 6.896551724137932e-06,
|
| 107 |
+
"loss": 0.8742586374282837,
|
| 108 |
+
"num_pairs": 28.0,
|
| 109 |
+
"step": 11
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"avg_label": 0.42196430265903473,
|
| 113 |
+
"epoch": 0.041666666666666664,
|
| 114 |
+
"grad_norm": 17.625,
|
| 115 |
+
"learning_rate": 7.586206896551724e-06,
|
| 116 |
+
"loss": 0.8603673875331879,
|
| 117 |
+
"num_pairs": 23.5,
|
| 118 |
+
"step": 12
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"avg_label": 0.44431088864803314,
|
| 122 |
+
"epoch": 0.04513888888888889,
|
| 123 |
+
"grad_norm": 1.3203125,
|
| 124 |
+
"learning_rate": 8.275862068965518e-06,
|
| 125 |
+
"loss": 0.028431319631636143,
|
| 126 |
+
"num_pairs": 28.0,
|
| 127 |
+
"step": 13
|
| 128 |
+
},
|
| 129 |
+
{
|
| 130 |
+
"avg_label": 0.3893229216337204,
|
| 131 |
+
"epoch": 0.04861111111111111,
|
| 132 |
+
"grad_norm": 10.0,
|
| 133 |
+
"learning_rate": 8.965517241379312e-06,
|
| 134 |
+
"loss": 0.251303069293499,
|
| 135 |
+
"num_pairs": 23.5,
|
| 136 |
+
"step": 14
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"avg_label": 0.43348929286003113,
|
| 140 |
+
"epoch": 0.052083333333333336,
|
| 141 |
+
"grad_norm": 3.796875,
|
| 142 |
+
"learning_rate": 9.655172413793105e-06,
|
| 143 |
+
"loss": 0.11253293231129646,
|
| 144 |
+
"num_pairs": 27.0,
|
| 145 |
+
"step": 15
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"avg_label": 0.439510241150856,
|
| 149 |
+
"epoch": 0.05555555555555555,
|
| 150 |
+
"grad_norm": 4.21875,
|
| 151 |
+
"learning_rate": 1.0344827586206898e-05,
|
| 152 |
+
"loss": 0.16489703953266144,
|
| 153 |
+
"num_pairs": 28.0,
|
| 154 |
+
"step": 16
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"avg_label": 0.4412849396467209,
|
| 158 |
+
"epoch": 0.059027777777777776,
|
| 159 |
+
"grad_norm": 4.84375,
|
| 160 |
+
"learning_rate": 1.103448275862069e-05,
|
| 161 |
+
"loss": 0.17035143449902534,
|
| 162 |
+
"num_pairs": 28.0,
|
| 163 |
+
"step": 17
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"avg_label": 0.4661668390035629,
|
| 167 |
+
"epoch": 0.0625,
|
| 168 |
+
"grad_norm": 3.1875,
|
| 169 |
+
"learning_rate": 1.1724137931034483e-05,
|
| 170 |
+
"loss": 0.10208164528012276,
|
| 171 |
+
"num_pairs": 28.0,
|
| 172 |
+
"step": 18
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"avg_label": 0.45625001192092896,
|
| 176 |
+
"epoch": 0.06597222222222222,
|
| 177 |
+
"grad_norm": 13.625,
|
| 178 |
+
"learning_rate": 1.2413793103448277e-05,
|
| 179 |
+
"loss": 0.5987264770083129,
|
| 180 |
+
"num_pairs": 28.0,
|
| 181 |
+
"step": 19
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"avg_label": 0.4458743929862976,
|
| 185 |
+
"epoch": 0.06944444444444445,
|
| 186 |
+
"grad_norm": 0.59375,
|
| 187 |
+
"learning_rate": 1.310344827586207e-05,
|
| 188 |
+
"loss": 0.03994415677152574,
|
| 189 |
+
"num_pairs": 28.0,
|
| 190 |
+
"step": 20
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"avg_label": 0.4499442130327225,
|
| 194 |
+
"epoch": 0.07291666666666667,
|
| 195 |
+
"grad_norm": 2.078125,
|
| 196 |
+
"learning_rate": 1.3793103448275863e-05,
|
| 197 |
+
"loss": 0.08950073830783367,
|
| 198 |
+
"num_pairs": 28.0,
|
| 199 |
+
"step": 21
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"avg_label": 0.4587053656578064,
|
| 203 |
+
"epoch": 0.0763888888888889,
|
| 204 |
+
"grad_norm": 9.0625,
|
| 205 |
+
"learning_rate": 1.4482758620689657e-05,
|
| 206 |
+
"loss": 0.5212011188268661,
|
| 207 |
+
"num_pairs": 28.0,
|
| 208 |
+
"step": 22
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"avg_label": 0.4502655118703842,
|
| 212 |
+
"epoch": 0.0798611111111111,
|
| 213 |
+
"grad_norm": 5.03125,
|
| 214 |
+
"learning_rate": 1.5172413793103448e-05,
|
| 215 |
+
"loss": 0.08295552420895547,
|
| 216 |
+
"num_pairs": 28.0,
|
| 217 |
+
"step": 23
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"avg_label": 0.4362068921327591,
|
| 221 |
+
"epoch": 0.08333333333333333,
|
| 222 |
+
"grad_norm": 1.65625,
|
| 223 |
+
"learning_rate": 1.586206896551724e-05,
|
| 224 |
+
"loss": 0.07221807166934013,
|
| 225 |
+
"num_pairs": 27.0,
|
| 226 |
+
"step": 24
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"avg_label": 0.4532342702150345,
|
| 230 |
+
"epoch": 0.08680555555555555,
|
| 231 |
+
"grad_norm": 1.140625,
|
| 232 |
+
"learning_rate": 1.6551724137931037e-05,
|
| 233 |
+
"loss": 0.03893738705664873,
|
| 234 |
+
"num_pairs": 28.0,
|
| 235 |
+
"step": 25
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"avg_label": 0.4947916716337204,
|
| 239 |
+
"epoch": 0.09027777777777778,
|
| 240 |
+
"grad_norm": 6.28125,
|
| 241 |
+
"learning_rate": 1.7241379310344828e-05,
|
| 242 |
+
"loss": 0.13335712999105453,
|
| 243 |
+
"num_pairs": 23.5,
|
| 244 |
+
"step": 26
|
| 245 |
+
},
|
| 246 |
+
{
|
| 247 |
+
"avg_label": 0.3125,
|
| 248 |
+
"epoch": 0.09375,
|
| 249 |
+
"grad_norm": 13.75,
|
| 250 |
+
"learning_rate": 1.7931034482758623e-05,
|
| 251 |
+
"loss": 0.23121396452188492,
|
| 252 |
+
"num_pairs": 17.5,
|
| 253 |
+
"step": 27
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"avg_label": 0.4624594449996948,
|
| 257 |
+
"epoch": 0.09722222222222222,
|
| 258 |
+
"grad_norm": 11.75,
|
| 259 |
+
"learning_rate": 1.8620689655172415e-05,
|
| 260 |
+
"loss": 0.4965571314096451,
|
| 261 |
+
"num_pairs": 28.0,
|
| 262 |
+
"step": 28
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"avg_label": 0.46101999282836914,
|
| 266 |
+
"epoch": 0.10069444444444445,
|
| 267 |
+
"grad_norm": 3.578125,
|
| 268 |
+
"learning_rate": 1.931034482758621e-05,
|
| 269 |
+
"loss": 0.16813608072698116,
|
| 270 |
+
"num_pairs": 28.0,
|
| 271 |
+
"step": 29
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"avg_label": 0.4742646962404251,
|
| 275 |
+
"epoch": 0.10416666666666667,
|
| 276 |
+
"grad_norm": 3.171875,
|
| 277 |
+
"learning_rate": 2e-05,
|
| 278 |
+
"loss": 0.1642257682979107,
|
| 279 |
+
"num_pairs": 22.0,
|
| 280 |
+
"step": 30
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"avg_label": 0.45412661135196686,
|
| 284 |
+
"epoch": 0.1076388888888889,
|
| 285 |
+
"grad_norm": 8.4375,
|
| 286 |
+
"learning_rate": 1.9999835072185805e-05,
|
| 287 |
+
"loss": 0.24545209854841232,
|
| 288 |
+
"num_pairs": 28.0,
|
| 289 |
+
"step": 31
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"avg_label": 0.45036764442920685,
|
| 293 |
+
"epoch": 0.1111111111111111,
|
| 294 |
+
"grad_norm": 3.234375,
|
| 295 |
+
"learning_rate": 1.999934029418346e-05,
|
| 296 |
+
"loss": 0.14530762657523155,
|
| 297 |
+
"num_pairs": 26.0,
|
| 298 |
+
"step": 32
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"avg_label": 0.4388917088508606,
|
| 302 |
+
"epoch": 0.11458333333333333,
|
| 303 |
+
"grad_norm": 2.890625,
|
| 304 |
+
"learning_rate": 1.9998515682313485e-05,
|
| 305 |
+
"loss": 0.10479701962321997,
|
| 306 |
+
"num_pairs": 28.0,
|
| 307 |
+
"step": 33
|
| 308 |
+
},
|
| 309 |
+
{
|
| 310 |
+
"avg_label": 0.43915343284606934,
|
| 311 |
+
"epoch": 0.11805555555555555,
|
| 312 |
+
"grad_norm": 7.875,
|
| 313 |
+
"learning_rate": 1.999736126377618e-05,
|
| 314 |
+
"loss": 0.3016491085290909,
|
| 315 |
+
"num_pairs": 28.0,
|
| 316 |
+
"step": 34
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"avg_label": 0.4497767984867096,
|
| 320 |
+
"epoch": 0.12152777777777778,
|
| 321 |
+
"grad_norm": 3.96875,
|
| 322 |
+
"learning_rate": 1.999587707665068e-05,
|
| 323 |
+
"loss": 0.15345897153019905,
|
| 324 |
+
"num_pairs": 28.0,
|
| 325 |
+
"step": 35
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"avg_label": 0.4572916626930237,
|
| 329 |
+
"epoch": 0.125,
|
| 330 |
+
"grad_norm": 1.3515625,
|
| 331 |
+
"learning_rate": 1.999406316989374e-05,
|
| 332 |
+
"loss": 0.04903313983231783,
|
| 333 |
+
"num_pairs": 28.0,
|
| 334 |
+
"step": 36
|
| 335 |
+
},
|
| 336 |
+
{
|
| 337 |
+
"avg_label": 0.45703125,
|
| 338 |
+
"epoch": 0.1284722222222222,
|
| 339 |
+
"grad_norm": 8.6875,
|
| 340 |
+
"learning_rate": 1.9991919603338088e-05,
|
| 341 |
+
"loss": 0.36730772256851196,
|
| 342 |
+
"num_pairs": 28.0,
|
| 343 |
+
"step": 37
|
| 344 |
+
},
|
| 345 |
+
{
|
| 346 |
+
"avg_label": 0.4379105120897293,
|
| 347 |
+
"epoch": 0.13194444444444445,
|
| 348 |
+
"grad_norm": 6.25,
|
| 349 |
+
"learning_rate": 1.998944644769048e-05,
|
| 350 |
+
"loss": 0.2889542682096362,
|
| 351 |
+
"num_pairs": 27.5,
|
| 352 |
+
"step": 38
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"avg_label": 0.4781250059604645,
|
| 356 |
+
"epoch": 0.13541666666666666,
|
| 357 |
+
"grad_norm": 10.0625,
|
| 358 |
+
"learning_rate": 1.9986643784529346e-05,
|
| 359 |
+
"loss": 0.416560098528862,
|
| 360 |
+
"num_pairs": 24.5,
|
| 361 |
+
"step": 39
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"avg_label": 0.4678819328546524,
|
| 365 |
+
"epoch": 0.1388888888888889,
|
| 366 |
+
"grad_norm": 4.96875,
|
| 367 |
+
"learning_rate": 1.9983511706302102e-05,
|
| 368 |
+
"loss": 0.1765335127711296,
|
| 369 |
+
"num_pairs": 28.0,
|
| 370 |
+
"step": 40
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"avg_label": 0.4556451737880707,
|
| 374 |
+
"epoch": 0.1423611111111111,
|
| 375 |
+
"grad_norm": 3.171875,
|
| 376 |
+
"learning_rate": 1.9980050316322118e-05,
|
| 377 |
+
"loss": 0.15112879872322083,
|
| 378 |
+
"num_pairs": 27.5,
|
| 379 |
+
"step": 41
|
| 380 |
+
},
|
| 381 |
+
{
|
| 382 |
+
"avg_label": 0.4666379243135452,
|
| 383 |
+
"epoch": 0.14583333333333334,
|
| 384 |
+
"grad_norm": 10.875,
|
| 385 |
+
"learning_rate": 1.997625972876529e-05,
|
| 386 |
+
"loss": 0.5556365996599197,
|
| 387 |
+
"num_pairs": 28.0,
|
| 388 |
+
"step": 42
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"avg_label": 0.45076756179332733,
|
| 392 |
+
"epoch": 0.14930555555555555,
|
| 393 |
+
"grad_norm": 2.40625,
|
| 394 |
+
"learning_rate": 1.997214006866628e-05,
|
| 395 |
+
"loss": 0.09973586304113269,
|
| 396 |
+
"num_pairs": 28.0,
|
| 397 |
+
"step": 43
|
| 398 |
+
},
|
| 399 |
+
{
|
| 400 |
+
"avg_label": 0.4405048042535782,
|
| 401 |
+
"epoch": 0.1527777777777778,
|
| 402 |
+
"grad_norm": 8.9375,
|
| 403 |
+
"learning_rate": 1.9967691471914392e-05,
|
| 404 |
+
"loss": 0.3806912750005722,
|
| 405 |
+
"num_pairs": 28.0,
|
| 406 |
+
"step": 44
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"avg_label": 0.4602416008710861,
|
| 410 |
+
"epoch": 0.15625,
|
| 411 |
+
"grad_norm": 4.8125,
|
| 412 |
+
"learning_rate": 1.99629140852491e-05,
|
| 413 |
+
"loss": 0.14190024323761463,
|
| 414 |
+
"num_pairs": 28.0,
|
| 415 |
+
"step": 45
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"avg_label": 0.44902269542217255,
|
| 419 |
+
"epoch": 0.1597222222222222,
|
| 420 |
+
"grad_norm": 1.890625,
|
| 421 |
+
"learning_rate": 1.9957808066255187e-05,
|
| 422 |
+
"loss": 0.07964974269270897,
|
| 423 |
+
"num_pairs": 28.0,
|
| 424 |
+
"step": 46
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"avg_label": 0.44250571727752686,
|
| 428 |
+
"epoch": 0.16319444444444445,
|
| 429 |
+
"grad_norm": 7.5625,
|
| 430 |
+
"learning_rate": 1.9952373583357566e-05,
|
| 431 |
+
"loss": 0.3196651563048363,
|
| 432 |
+
"num_pairs": 27.5,
|
| 433 |
+
"step": 47
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"avg_label": 0.45804399251937866,
|
| 437 |
+
"epoch": 0.16666666666666666,
|
| 438 |
+
"grad_norm": 3.03125,
|
| 439 |
+
"learning_rate": 1.994661081581571e-05,
|
| 440 |
+
"loss": 0.1772424913942814,
|
| 441 |
+
"num_pairs": 28.0,
|
| 442 |
+
"step": 48
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"avg_label": 0.44343800842761993,
|
| 446 |
+
"epoch": 0.1701388888888889,
|
| 447 |
+
"grad_norm": 0.7734375,
|
| 448 |
+
"learning_rate": 1.9940519953717762e-05,
|
| 449 |
+
"loss": 0.03773397207260132,
|
| 450 |
+
"num_pairs": 27.5,
|
| 451 |
+
"step": 49
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"avg_label": 0.4277420938014984,
|
| 455 |
+
"epoch": 0.1736111111111111,
|
| 456 |
+
"grad_norm": 0.7578125,
|
| 457 |
+
"learning_rate": 1.993410119797422e-05,
|
| 458 |
+
"loss": 0.017126482212916017,
|
| 459 |
+
"num_pairs": 27.5,
|
| 460 |
+
"step": 50
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"avg_label": 0.4761904776096344,
|
| 464 |
+
"epoch": 0.17708333333333334,
|
| 465 |
+
"grad_norm": 2.265625,
|
| 466 |
+
"learning_rate": 1.9927354760311365e-05,
|
| 467 |
+
"loss": 0.0958416610956192,
|
| 468 |
+
"num_pairs": 24.0,
|
| 469 |
+
"step": 51
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"avg_label": 0.45110294222831726,
|
| 473 |
+
"epoch": 0.18055555555555555,
|
| 474 |
+
"grad_norm": 3.859375,
|
| 475 |
+
"learning_rate": 1.992028086326424e-05,
|
| 476 |
+
"loss": 0.176346430554986,
|
| 477 |
+
"num_pairs": 28.0,
|
| 478 |
+
"step": 52
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"avg_label": 0.4557291716337204,
|
| 482 |
+
"epoch": 0.1840277777777778,
|
| 483 |
+
"grad_norm": 1.484375,
|
| 484 |
+
"learning_rate": 1.991287974016932e-05,
|
| 485 |
+
"loss": 0.06965811923146248,
|
| 486 |
+
"num_pairs": 28.0,
|
| 487 |
+
"step": 53
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"avg_label": 0.441498339176178,
|
| 491 |
+
"epoch": 0.1875,
|
| 492 |
+
"grad_norm": 1.1328125,
|
| 493 |
+
"learning_rate": 1.9905151635156813e-05,
|
| 494 |
+
"loss": 0.04886321909725666,
|
| 495 |
+
"num_pairs": 28.0,
|
| 496 |
+
"step": 54
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"avg_label": 0.4579603523015976,
|
| 500 |
+
"epoch": 0.1909722222222222,
|
| 501 |
+
"grad_norm": 0.79296875,
|
| 502 |
+
"learning_rate": 1.9897096803142616e-05,
|
| 503 |
+
"loss": 0.02616998180747032,
|
| 504 |
+
"num_pairs": 28.0,
|
| 505 |
+
"step": 55
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"avg_label": 0.45521390438079834,
|
| 509 |
+
"epoch": 0.19444444444444445,
|
| 510 |
+
"grad_norm": 0.48046875,
|
| 511 |
+
"learning_rate": 1.988871550981989e-05,
|
| 512 |
+
"loss": 0.04496368020772934,
|
| 513 |
+
"num_pairs": 28.0,
|
| 514 |
+
"step": 56
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"avg_label": 0.46743176877498627,
|
| 518 |
+
"epoch": 0.19791666666666666,
|
| 519 |
+
"grad_norm": 1.953125,
|
| 520 |
+
"learning_rate": 1.988000803165032e-05,
|
| 521 |
+
"loss": 0.08031387068331242,
|
| 522 |
+
"num_pairs": 27.0,
|
| 523 |
+
"step": 57
|
| 524 |
+
},
|
| 525 |
+
{
|
| 526 |
+
"avg_label": 0.39706501364707947,
|
| 527 |
+
"epoch": 0.2013888888888889,
|
| 528 |
+
"grad_norm": 0.765625,
|
| 529 |
+
"learning_rate": 1.9870974655854974e-05,
|
| 530 |
+
"loss": 0.05276821367442608,
|
| 531 |
+
"num_pairs": 28.0,
|
| 532 |
+
"step": 58
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"avg_label": 0.4635416716337204,
|
| 536 |
+
"epoch": 0.2048611111111111,
|
| 537 |
+
"grad_norm": 0.345703125,
|
| 538 |
+
"learning_rate": 1.9861615680404833e-05,
|
| 539 |
+
"loss": 0.02046751804300584,
|
| 540 |
+
"num_pairs": 28.0,
|
| 541 |
+
"step": 59
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"avg_label": 0.4392857253551483,
|
| 545 |
+
"epoch": 0.20833333333333334,
|
| 546 |
+
"grad_norm": 1.59375,
|
| 547 |
+
"learning_rate": 1.985193141401097e-05,
|
| 548 |
+
"loss": 0.03745671547949314,
|
| 549 |
+
"num_pairs": 28.0,
|
| 550 |
+
"step": 60
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"avg_label": 0.430803582072258,
|
| 554 |
+
"epoch": 0.21180555555555555,
|
| 555 |
+
"grad_norm": 2.34375,
|
| 556 |
+
"learning_rate": 1.9841922176114366e-05,
|
| 557 |
+
"loss": 0.04785962332971394,
|
| 558 |
+
"num_pairs": 21.5,
|
| 559 |
+
"step": 61
|
| 560 |
+
},
|
| 561 |
+
{
|
| 562 |
+
"avg_label": 0.4626736044883728,
|
| 563 |
+
"epoch": 0.2152777777777778,
|
| 564 |
+
"grad_norm": 0.61328125,
|
| 565 |
+
"learning_rate": 1.9831588296875367e-05,
|
| 566 |
+
"loss": 0.021605145651847124,
|
| 567 |
+
"num_pairs": 28.0,
|
| 568 |
+
"step": 62
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"avg_label": 0.4780505895614624,
|
| 572 |
+
"epoch": 0.21875,
|
| 573 |
+
"grad_norm": 1.609375,
|
| 574 |
+
"learning_rate": 1.982093011716279e-05,
|
| 575 |
+
"loss": 0.08873376995325089,
|
| 576 |
+
"num_pairs": 28.0,
|
| 577 |
+
"step": 63
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"avg_label": 0.4489087462425232,
|
| 581 |
+
"epoch": 0.2222222222222222,
|
| 582 |
+
"grad_norm": 1.65625,
|
| 583 |
+
"learning_rate": 1.9809947988542696e-05,
|
| 584 |
+
"loss": 0.0287565803155303,
|
| 585 |
+
"num_pairs": 27.5,
|
| 586 |
+
"step": 64
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"avg_label": 0.4442349076271057,
|
| 590 |
+
"epoch": 0.22569444444444445,
|
| 591 |
+
"grad_norm": 0.82421875,
|
| 592 |
+
"learning_rate": 1.979864227326678e-05,
|
| 593 |
+
"loss": 0.041462352965027094,
|
| 594 |
+
"num_pairs": 28.0,
|
| 595 |
+
"step": 65
|
| 596 |
+
},
|
| 597 |
+
{
|
| 598 |
+
"avg_label": 0.5274767875671387,
|
| 599 |
+
"epoch": 0.22916666666666666,
|
| 600 |
+
"grad_norm": 0.26171875,
|
| 601 |
+
"learning_rate": 1.9787013344260422e-05,
|
| 602 |
+
"loss": 0.006527273333631456,
|
| 603 |
+
"num_pairs": 26.0,
|
| 604 |
+
"step": 66
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"avg_label": 0.45734797418117523,
|
| 608 |
+
"epoch": 0.2326388888888889,
|
| 609 |
+
"grad_norm": 7.21875,
|
| 610 |
+
"learning_rate": 1.9775061585110387e-05,
|
| 611 |
+
"loss": 0.33710330724716187,
|
| 612 |
+
"num_pairs": 28.0,
|
| 613 |
+
"step": 67
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"avg_label": 0.44953545928001404,
|
| 617 |
+
"epoch": 0.2361111111111111,
|
| 618 |
+
"grad_norm": 1.1796875,
|
| 619 |
+
"learning_rate": 1.976278739005218e-05,
|
| 620 |
+
"loss": 0.05175479780882597,
|
| 621 |
+
"num_pairs": 28.0,
|
| 622 |
+
"step": 68
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"avg_label": 0.4675245136022568,
|
| 626 |
+
"epoch": 0.23958333333333334,
|
| 627 |
+
"grad_norm": 3.859375,
|
| 628 |
+
"learning_rate": 1.9750191163957042e-05,
|
| 629 |
+
"loss": 0.15232571214437485,
|
| 630 |
+
"num_pairs": 28.0,
|
| 631 |
+
"step": 69
|
| 632 |
+
},
|
| 633 |
+
{
|
| 634 |
+
"avg_label": 0.45781250298023224,
|
| 635 |
+
"epoch": 0.24305555555555555,
|
| 636 |
+
"grad_norm": 1.7890625,
|
| 637 |
+
"learning_rate": 1.9737273322318565e-05,
|
| 638 |
+
"loss": 0.09527681209146976,
|
| 639 |
+
"num_pairs": 28.0,
|
| 640 |
+
"step": 70
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"avg_label": 0.4518994987010956,
|
| 644 |
+
"epoch": 0.2465277777777778,
|
| 645 |
+
"grad_norm": 0.703125,
|
| 646 |
+
"learning_rate": 1.972403429123904e-05,
|
| 647 |
+
"loss": 0.035001321870367974,
|
| 648 |
+
"num_pairs": 28.0,
|
| 649 |
+
"step": 71
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"avg_label": 0.4765625,
|
| 653 |
+
"epoch": 0.25,
|
| 654 |
+
"grad_norm": 0.75,
|
| 655 |
+
"learning_rate": 1.971047450741535e-05,
|
| 656 |
+
"loss": 0.026095230132341385,
|
| 657 |
+
"num_pairs": 22.0,
|
| 658 |
+
"step": 72
|
| 659 |
+
},
|
| 660 |
+
{
|
| 661 |
+
"avg_label": 0.5010775923728943,
|
| 662 |
+
"epoch": 0.2534722222222222,
|
| 663 |
+
"grad_norm": 0.58984375,
|
| 664 |
+
"learning_rate": 1.9696594418124598e-05,
|
| 665 |
+
"loss": 0.028016670839861035,
|
| 666 |
+
"num_pairs": 23.5,
|
| 667 |
+
"step": 73
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"avg_label": 0.6005434691905975,
|
| 671 |
+
"epoch": 0.2569444444444444,
|
| 672 |
+
"grad_norm": 0.62109375,
|
| 673 |
+
"learning_rate": 1.9682394481209338e-05,
|
| 674 |
+
"loss": 0.017244269140064716,
|
| 675 |
+
"num_pairs": 20.0,
|
| 676 |
+
"step": 74
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"avg_label": 0.44895362854003906,
|
| 680 |
+
"epoch": 0.2604166666666667,
|
| 681 |
+
"grad_norm": 2.65625,
|
| 682 |
+
"learning_rate": 1.966787516506249e-05,
|
| 683 |
+
"loss": 0.13350838515907526,
|
| 684 |
+
"num_pairs": 27.5,
|
| 685 |
+
"step": 75
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"avg_label": 0.44657258689403534,
|
| 689 |
+
"epoch": 0.2638888888888889,
|
| 690 |
+
"grad_norm": 2.6875,
|
| 691 |
+
"learning_rate": 1.9653036948611864e-05,
|
| 692 |
+
"loss": 0.13598253019154072,
|
| 693 |
+
"num_pairs": 24.5,
|
| 694 |
+
"step": 76
|
| 695 |
+
},
|
| 696 |
+
{
|
| 697 |
+
"avg_label": 0.4062500149011612,
|
| 698 |
+
"epoch": 0.2673611111111111,
|
| 699 |
+
"grad_norm": 1.6328125,
|
| 700 |
+
"learning_rate": 1.9637880321304387e-05,
|
| 701 |
+
"loss": 0.09597307164222002,
|
| 702 |
+
"num_pairs": 21.5,
|
| 703 |
+
"step": 77
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"avg_label": 0.43358150124549866,
|
| 707 |
+
"epoch": 0.2708333333333333,
|
| 708 |
+
"grad_norm": 1.5859375,
|
| 709 |
+
"learning_rate": 1.962240578308993e-05,
|
| 710 |
+
"loss": 0.02830888982862234,
|
| 711 |
+
"num_pairs": 28.0,
|
| 712 |
+
"step": 78
|
| 713 |
+
},
|
| 714 |
+
{
|
| 715 |
+
"avg_label": 0.4296875,
|
| 716 |
+
"epoch": 0.2743055555555556,
|
| 717 |
+
"grad_norm": 0.5703125,
|
| 718 |
+
"learning_rate": 1.9606613844404853e-05,
|
| 719 |
+
"loss": 0.0342620718292892,
|
| 720 |
+
"num_pairs": 27.5,
|
| 721 |
+
"step": 79
|
| 722 |
+
},
|
| 723 |
+
{
|
| 724 |
+
"avg_label": 0.44969919323921204,
|
| 725 |
+
"epoch": 0.2777777777777778,
|
| 726 |
+
"grad_norm": 1.140625,
|
| 727 |
+
"learning_rate": 1.9590505026155146e-05,
|
| 728 |
+
"loss": 0.047657732968218625,
|
| 729 |
+
"num_pairs": 28.0,
|
| 730 |
+
"step": 80
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"avg_label": 0.44572368264198303,
|
| 734 |
+
"epoch": 0.28125,
|
| 735 |
+
"grad_norm": 0.98828125,
|
| 736 |
+
"learning_rate": 1.9574079859699236e-05,
|
| 737 |
+
"loss": 0.048789044842123985,
|
| 738 |
+
"num_pairs": 28.0,
|
| 739 |
+
"step": 81
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"avg_label": 0.39816810190677643,
|
| 743 |
+
"epoch": 0.2847222222222222,
|
| 744 |
+
"grad_norm": 1.140625,
|
| 745 |
+
"learning_rate": 1.955733888683049e-05,
|
| 746 |
+
"loss": 0.053058773279190063,
|
| 747 |
+
"num_pairs": 27.5,
|
| 748 |
+
"step": 82
|
| 749 |
+
},
|
| 750 |
+
{
|
| 751 |
+
"avg_label": 0.43917112052440643,
|
| 752 |
+
"epoch": 0.2881944444444444,
|
| 753 |
+
"grad_norm": 0.91015625,
|
| 754 |
+
"learning_rate": 1.9540282659759317e-05,
|
| 755 |
+
"loss": 0.05478274542838335,
|
| 756 |
+
"num_pairs": 27.5,
|
| 757 |
+
"step": 83
|
| 758 |
+
},
|
| 759 |
+
{
|
| 760 |
+
"avg_label": 0.4709821492433548,
|
| 761 |
+
"epoch": 0.2916666666666667,
|
| 762 |
+
"grad_norm": 0.228515625,
|
| 763 |
+
"learning_rate": 1.9522911741094966e-05,
|
| 764 |
+
"loss": 0.011658322997391224,
|
| 765 |
+
"num_pairs": 28.0,
|
| 766 |
+
"step": 84
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"avg_label": 0.4544480890035629,
|
| 770 |
+
"epoch": 0.2951388888888889,
|
| 771 |
+
"grad_norm": 11.3125,
|
| 772 |
+
"learning_rate": 1.9505226703826973e-05,
|
| 773 |
+
"loss": 0.523826252669096,
|
| 774 |
+
"num_pairs": 28.0,
|
| 775 |
+
"step": 85
|
| 776 |
+
},
|
| 777 |
+
{
|
| 778 |
+
"avg_label": 0.4711538553237915,
|
| 779 |
+
"epoch": 0.2986111111111111,
|
| 780 |
+
"grad_norm": 0.80078125,
|
| 781 |
+
"learning_rate": 1.948722813130624e-05,
|
| 782 |
+
"loss": 0.023038258543238044,
|
| 783 |
+
"num_pairs": 27.5,
|
| 784 |
+
"step": 86
|
| 785 |
+
},
|
| 786 |
+
{
|
| 787 |
+
"avg_label": 0.4539930522441864,
|
| 788 |
+
"epoch": 0.3020833333333333,
|
| 789 |
+
"grad_norm": 1.6328125,
|
| 790 |
+
"learning_rate": 1.9468916617225814e-05,
|
| 791 |
+
"loss": 0.058613574132323265,
|
| 792 |
+
"num_pairs": 28.0,
|
| 793 |
+
"step": 87
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"avg_label": 0.4584091007709503,
|
| 797 |
+
"epoch": 0.3055555555555556,
|
| 798 |
+
"grad_norm": 0.388671875,
|
| 799 |
+
"learning_rate": 1.9450292765601287e-05,
|
| 800 |
+
"loss": 0.016918038250878453,
|
| 801 |
+
"num_pairs": 28.0,
|
| 802 |
+
"step": 88
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"avg_label": 0.4232954829931259,
|
| 806 |
+
"epoch": 0.3090277777777778,
|
| 807 |
+
"grad_norm": 1.3359375,
|
| 808 |
+
"learning_rate": 1.94313571907509e-05,
|
| 809 |
+
"loss": 0.060430300422012806,
|
| 810 |
+
"num_pairs": 28.0,
|
| 811 |
+
"step": 89
|
| 812 |
+
},
|
| 813 |
+
{
|
| 814 |
+
"avg_label": 0.5833333134651184,
|
| 815 |
+
"epoch": 0.3125,
|
| 816 |
+
"grad_norm": 4.125,
|
| 817 |
+
"learning_rate": 1.941211051727524e-05,
|
| 818 |
+
"loss": 0.060059962212108076,
|
| 819 |
+
"num_pairs": 20.5,
|
| 820 |
+
"step": 90
|
| 821 |
+
},
|
| 822 |
+
{
|
| 823 |
+
"avg_label": 0.45364584028720856,
|
| 824 |
+
"epoch": 0.3159722222222222,
|
| 825 |
+
"grad_norm": 1.4375,
|
| 826 |
+
"learning_rate": 1.939255338003666e-05,
|
| 827 |
+
"loss": 0.04583445412572473,
|
| 828 |
+
"num_pairs": 28.0,
|
| 829 |
+
"step": 91
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"avg_label": 0.4437500089406967,
|
| 833 |
+
"epoch": 0.3194444444444444,
|
| 834 |
+
"grad_norm": 0.26171875,
|
| 835 |
+
"learning_rate": 1.937268642413835e-05,
|
| 836 |
+
"loss": 0.00882079591974616,
|
| 837 |
+
"num_pairs": 27.0,
|
| 838 |
+
"step": 92
|
| 839 |
+
},
|
| 840 |
+
{
|
| 841 |
+
"avg_label": 0.4508301317691803,
|
| 842 |
+
"epoch": 0.3229166666666667,
|
| 843 |
+
"grad_norm": 0.51953125,
|
| 844 |
+
"learning_rate": 1.9352510304903017e-05,
|
| 845 |
+
"loss": 0.03392870319657959,
|
| 846 |
+
"num_pairs": 27.5,
|
| 847 |
+
"step": 93
|
| 848 |
+
},
|
| 849 |
+
{
|
| 850 |
+
"avg_label": 0.44549499452114105,
|
| 851 |
+
"epoch": 0.3263888888888889,
|
| 852 |
+
"grad_norm": 1.4375,
|
| 853 |
+
"learning_rate": 1.9332025687851325e-05,
|
| 854 |
+
"loss": 0.04178585158661008,
|
| 855 |
+
"num_pairs": 28.0,
|
| 856 |
+
"step": 94
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"avg_label": 0.47159090638160706,
|
| 860 |
+
"epoch": 0.3298611111111111,
|
| 861 |
+
"grad_norm": 0.92578125,
|
| 862 |
+
"learning_rate": 1.931123324867989e-05,
|
| 863 |
+
"loss": 0.028863655403256416,
|
| 864 |
+
"num_pairs": 28.0,
|
| 865 |
+
"step": 95
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"avg_label": 0.4557291716337204,
|
| 869 |
+
"epoch": 0.3333333333333333,
|
| 870 |
+
"grad_norm": 2.671875,
|
| 871 |
+
"learning_rate": 1.929013367323902e-05,
|
| 872 |
+
"loss": 0.1290583610534668,
|
| 873 |
+
"num_pairs": 28.0,
|
| 874 |
+
"step": 96
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"avg_label": 0.44440363347530365,
|
| 878 |
+
"epoch": 0.3368055555555556,
|
| 879 |
+
"grad_norm": 0.75390625,
|
| 880 |
+
"learning_rate": 1.926872765751009e-05,
|
| 881 |
+
"loss": 0.01571450149640441,
|
| 882 |
+
"num_pairs": 28.0,
|
| 883 |
+
"step": 97
|
| 884 |
+
},
|
| 885 |
+
{
|
| 886 |
+
"avg_label": 0.5104166716337204,
|
| 887 |
+
"epoch": 0.3402777777777778,
|
| 888 |
+
"grad_norm": 5.9375,
|
| 889 |
+
"learning_rate": 1.9247015907582574e-05,
|
| 890 |
+
"loss": 0.2692837491631508,
|
| 891 |
+
"num_pairs": 27.5,
|
| 892 |
+
"step": 98
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"avg_label": 0.4975000023841858,
|
| 896 |
+
"epoch": 0.34375,
|
| 897 |
+
"grad_norm": 0.462890625,
|
| 898 |
+
"learning_rate": 1.9224999139630766e-05,
|
| 899 |
+
"loss": 0.02842018473893404,
|
| 900 |
+
"num_pairs": 24.0,
|
| 901 |
+
"step": 99
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"avg_label": 0.46345899999141693,
|
| 905 |
+
"epoch": 0.3472222222222222,
|
| 906 |
+
"grad_norm": 3.578125,
|
| 907 |
+
"learning_rate": 1.920267807989015e-05,
|
| 908 |
+
"loss": 0.12233215570449829,
|
| 909 |
+
"num_pairs": 25.5,
|
| 910 |
+
"step": 100
|
| 911 |
+
},
|
| 912 |
+
{
|
| 913 |
+
"avg_label": 0.4338157922029495,
|
| 914 |
+
"epoch": 0.3506944444444444,
|
| 915 |
+
"grad_norm": 0.275390625,
|
| 916 |
+
"learning_rate": 1.918005346463344e-05,
|
| 917 |
+
"loss": 0.004982678568921983,
|
| 918 |
+
"num_pairs": 27.5,
|
| 919 |
+
"step": 101
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"avg_label": 0.4490041136741638,
|
| 923 |
+
"epoch": 0.3541666666666667,
|
| 924 |
+
"grad_norm": 0.427734375,
|
| 925 |
+
"learning_rate": 1.9157126040146307e-05,
|
| 926 |
+
"loss": 0.01375208212994039,
|
| 927 |
+
"num_pairs": 27.5,
|
| 928 |
+
"step": 102
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"avg_label": 0.4447794109582901,
|
| 932 |
+
"epoch": 0.3576388888888889,
|
| 933 |
+
"grad_norm": 5.125,
|
| 934 |
+
"learning_rate": 1.9133896562702746e-05,
|
| 935 |
+
"loss": 0.20500781387090683,
|
| 936 |
+
"num_pairs": 27.5,
|
| 937 |
+
"step": 103
|
| 938 |
+
},
|
| 939 |
+
{
|
| 940 |
+
"avg_label": 0.45121754705905914,
|
| 941 |
+
"epoch": 0.3611111111111111,
|
| 942 |
+
"grad_norm": 1.015625,
|
| 943 |
+
"learning_rate": 1.911036579854016e-05,
|
| 944 |
+
"loss": 0.024166605668142438,
|
| 945 |
+
"num_pairs": 28.0,
|
| 946 |
+
"step": 104
|
| 947 |
+
},
|
| 948 |
+
{
|
| 949 |
+
"avg_label": 0.5431034564971924,
|
| 950 |
+
"epoch": 0.3645833333333333,
|
| 951 |
+
"grad_norm": 0.87109375,
|
| 952 |
+
"learning_rate": 1.9086534523834032e-05,
|
| 953 |
+
"loss": 0.07264208607375622,
|
| 954 |
+
"num_pairs": 21.5,
|
| 955 |
+
"step": 105
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"avg_label": 0.5112500041723251,
|
| 959 |
+
"epoch": 0.3680555555555556,
|
| 960 |
+
"grad_norm": 0.5,
|
| 961 |
+
"learning_rate": 1.906240352467238e-05,
|
| 962 |
+
"loss": 0.027374916709959507,
|
| 963 |
+
"num_pairs": 24.5,
|
| 964 |
+
"step": 106
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"avg_label": 0.4663461595773697,
|
| 968 |
+
"epoch": 0.3715277777777778,
|
| 969 |
+
"grad_norm": 0.88671875,
|
| 970 |
+
"learning_rate": 1.9037973597029796e-05,
|
| 971 |
+
"loss": 0.009546086774207652,
|
| 972 |
+
"num_pairs": 24.0,
|
| 973 |
+
"step": 107
|
| 974 |
+
},
|
| 975 |
+
{
|
| 976 |
+
"avg_label": 0.45906251668930054,
|
| 977 |
+
"epoch": 0.375,
|
| 978 |
+
"grad_norm": 0.74609375,
|
| 979 |
+
"learning_rate": 1.901324554674119e-05,
|
| 980 |
+
"loss": 0.021506985649466515,
|
| 981 |
+
"num_pairs": 27.5,
|
| 982 |
+
"step": 108
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"avg_label": 0.5089285671710968,
|
| 986 |
+
"epoch": 0.3784722222222222,
|
| 987 |
+
"grad_norm": 0.82421875,
|
| 988 |
+
"learning_rate": 1.8988220189475216e-05,
|
| 989 |
+
"loss": 0.021596498914732365,
|
| 990 |
+
"num_pairs": 24.5,
|
| 991 |
+
"step": 109
|
| 992 |
+
},
|
| 993 |
+
{
|
| 994 |
+
"avg_label": 0.45194804668426514,
|
| 995 |
+
"epoch": 0.3819444444444444,
|
| 996 |
+
"grad_norm": 0.44921875,
|
| 997 |
+
"learning_rate": 1.896289835070737e-05,
|
| 998 |
+
"loss": 0.044558977242559195,
|
| 999 |
+
"num_pairs": 28.0,
|
| 1000 |
+
"step": 110
|
| 1001 |
+
},
|
| 1002 |
+
{
|
| 1003 |
+
"avg_label": 0.548076942563057,
|
| 1004 |
+
"epoch": 0.3854166666666667,
|
| 1005 |
+
"grad_norm": 5.34375,
|
| 1006 |
+
"learning_rate": 1.893728086569276e-05,
|
| 1007 |
+
"loss": 0.13392094970913604,
|
| 1008 |
+
"num_pairs": 21.5,
|
| 1009 |
+
"step": 111
|
| 1010 |
+
},
|
| 1011 |
+
{
|
| 1012 |
+
"avg_label": 0.4564732164144516,
|
| 1013 |
+
"epoch": 0.3888888888888889,
|
| 1014 |
+
"grad_norm": 1.796875,
|
| 1015 |
+
"learning_rate": 1.891136857943854e-05,
|
| 1016 |
+
"loss": 0.07770301774144173,
|
| 1017 |
+
"num_pairs": 28.0,
|
| 1018 |
+
"step": 112
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"avg_label": 0.4529440253973007,
|
| 1022 |
+
"epoch": 0.3923611111111111,
|
| 1023 |
+
"grad_norm": 0.609375,
|
| 1024 |
+
"learning_rate": 1.8885162346676063e-05,
|
| 1025 |
+
"loss": 0.03013956267386675,
|
| 1026 |
+
"num_pairs": 28.0,
|
| 1027 |
+
"step": 113
|
| 1028 |
+
},
|
| 1029 |
+
{
|
| 1030 |
+
"avg_label": 0.45974941551685333,
|
| 1031 |
+
"epoch": 0.3958333333333333,
|
| 1032 |
+
"grad_norm": 2.328125,
|
| 1033 |
+
"learning_rate": 1.8858663031832665e-05,
|
| 1034 |
+
"loss": 0.08871974685462192,
|
| 1035 |
+
"num_pairs": 28.0,
|
| 1036 |
+
"step": 114
|
| 1037 |
+
},
|
| 1038 |
+
{
|
| 1039 |
+
"avg_label": 0.45034562051296234,
|
| 1040 |
+
"epoch": 0.3993055555555556,
|
| 1041 |
+
"grad_norm": 0.1533203125,
|
| 1042 |
+
"learning_rate": 1.8831871509003164e-05,
|
| 1043 |
+
"loss": 0.002237841486930847,
|
| 1044 |
+
"num_pairs": 28.0,
|
| 1045 |
+
"step": 115
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"avg_label": 0.46330973505973816,
|
| 1049 |
+
"epoch": 0.4027777777777778,
|
| 1050 |
+
"grad_norm": 1.4140625,
|
| 1051 |
+
"learning_rate": 1.8804788661921012e-05,
|
| 1052 |
+
"loss": 0.034147982485592365,
|
| 1053 |
+
"num_pairs": 28.0,
|
| 1054 |
+
"step": 116
|
| 1055 |
+
},
|
| 1056 |
+
{
|
| 1057 |
+
"avg_label": 0.4184027910232544,
|
| 1058 |
+
"epoch": 0.40625,
|
| 1059 |
+
"grad_norm": 0.66015625,
|
| 1060 |
+
"learning_rate": 1.877741538392917e-05,
|
| 1061 |
+
"loss": 0.032842551125213504,
|
| 1062 |
+
"num_pairs": 28.0,
|
| 1063 |
+
"step": 117
|
| 1064 |
+
},
|
| 1065 |
+
{
|
| 1066 |
+
"avg_label": 0.44732142984867096,
|
| 1067 |
+
"epoch": 0.4097222222222222,
|
| 1068 |
+
"grad_norm": 0.50390625,
|
| 1069 |
+
"learning_rate": 1.8749752577950614e-05,
|
| 1070 |
+
"loss": 0.022164800204336643,
|
| 1071 |
+
"num_pairs": 23.5,
|
| 1072 |
+
"step": 118
|
| 1073 |
+
},
|
| 1074 |
+
{
|
| 1075 |
+
"avg_label": 0.4625000059604645,
|
| 1076 |
+
"epoch": 0.4131944444444444,
|
| 1077 |
+
"grad_norm": 2.46875,
|
| 1078 |
+
"learning_rate": 1.8721801156458573e-05,
|
| 1079 |
+
"loss": 0.13851431757211685,
|
| 1080 |
+
"num_pairs": 22.0,
|
| 1081 |
+
"step": 119
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"avg_label": 0.46339286863803864,
|
| 1085 |
+
"epoch": 0.4166666666666667,
|
| 1086 |
+
"grad_norm": 1.4140625,
|
| 1087 |
+
"learning_rate": 1.869356204144642e-05,
|
| 1088 |
+
"loss": 0.03970568906515837,
|
| 1089 |
+
"num_pairs": 26.0,
|
| 1090 |
+
"step": 120
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"avg_label": 0.46469351649284363,
|
| 1094 |
+
"epoch": 0.4201388888888889,
|
| 1095 |
+
"grad_norm": 1.4765625,
|
| 1096 |
+
"learning_rate": 1.866503616439725e-05,
|
| 1097 |
+
"loss": 0.06737381592392921,
|
| 1098 |
+
"num_pairs": 27.5,
|
| 1099 |
+
"step": 121
|
| 1100 |
+
},
|
| 1101 |
+
{
|
| 1102 |
+
"avg_label": 0.44958437979221344,
|
| 1103 |
+
"epoch": 0.4236111111111111,
|
| 1104 |
+
"grad_norm": 0.796875,
|
| 1105 |
+
"learning_rate": 1.8636224466253177e-05,
|
| 1106 |
+
"loss": 0.03086982760578394,
|
| 1107 |
+
"num_pairs": 28.0,
|
| 1108 |
+
"step": 122
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"avg_label": 0.4493534415960312,
|
| 1112 |
+
"epoch": 0.4270833333333333,
|
| 1113 |
+
"grad_norm": 1.0078125,
|
| 1114 |
+
"learning_rate": 1.860712789738428e-05,
|
| 1115 |
+
"loss": 0.03511205594986677,
|
| 1116 |
+
"num_pairs": 28.0,
|
| 1117 |
+
"step": 123
|
| 1118 |
+
},
|
| 1119 |
+
{
|
| 1120 |
+
"avg_label": 0.44804346561431885,
|
| 1121 |
+
"epoch": 0.4305555555555556,
|
| 1122 |
+
"grad_norm": 0.78125,
|
| 1123 |
+
"learning_rate": 1.857774741755726e-05,
|
| 1124 |
+
"loss": 0.056939804926514626,
|
| 1125 |
+
"num_pairs": 27.0,
|
| 1126 |
+
"step": 124
|
| 1127 |
+
},
|
| 1128 |
+
{
|
| 1129 |
+
"avg_label": 0.4523809552192688,
|
| 1130 |
+
"epoch": 0.4340277777777778,
|
| 1131 |
+
"grad_norm": 2.375,
|
| 1132 |
+
"learning_rate": 1.854808399590378e-05,
|
| 1133 |
+
"loss": 0.09799622371792793,
|
| 1134 |
+
"num_pairs": 28.0,
|
| 1135 |
+
"step": 125
|
| 1136 |
+
},
|
| 1137 |
+
{
|
| 1138 |
+
"avg_label": 0.4740624874830246,
|
| 1139 |
+
"epoch": 0.4375,
|
| 1140 |
+
"grad_norm": 1.625,
|
| 1141 |
+
"learning_rate": 1.8518138610888505e-05,
|
| 1142 |
+
"loss": 0.04024723544716835,
|
| 1143 |
+
"num_pairs": 28.0,
|
| 1144 |
+
"step": 126
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"avg_label": 0.4765625,
|
| 1148 |
+
"epoch": 0.4409722222222222,
|
| 1149 |
+
"grad_norm": 4.71875,
|
| 1150 |
+
"learning_rate": 1.8487912250276805e-05,
|
| 1151 |
+
"loss": 0.23970045894384384,
|
| 1152 |
+
"num_pairs": 24.0,
|
| 1153 |
+
"step": 127
|
| 1154 |
+
},
|
| 1155 |
+
{
|
| 1156 |
+
"avg_label": 0.4485294073820114,
|
| 1157 |
+
"epoch": 0.4444444444444444,
|
| 1158 |
+
"grad_norm": 0.87890625,
|
| 1159 |
+
"learning_rate": 1.8457405911102202e-05,
|
| 1160 |
+
"loss": 0.024124122224748135,
|
| 1161 |
+
"num_pairs": 24.5,
|
| 1162 |
+
"step": 128
|
| 1163 |
+
},
|
| 1164 |
+
{
|
| 1165 |
+
"avg_label": 0.470486119389534,
|
| 1166 |
+
"epoch": 0.4479166666666667,
|
| 1167 |
+
"grad_norm": 7.21875,
|
| 1168 |
+
"learning_rate": 1.8426620599633464e-05,
|
| 1169 |
+
"loss": 0.2318284586071968,
|
| 1170 |
+
"num_pairs": 22.0,
|
| 1171 |
+
"step": 129
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"avg_label": 0.4488866627216339,
|
| 1175 |
+
"epoch": 0.4513888888888889,
|
| 1176 |
+
"grad_norm": 0.6953125,
|
| 1177 |
+
"learning_rate": 1.8395557331341413e-05,
|
| 1178 |
+
"loss": 0.015422451309859753,
|
| 1179 |
+
"num_pairs": 28.0,
|
| 1180 |
+
"step": 130
|
| 1181 |
+
},
|
| 1182 |
+
{
|
| 1183 |
+
"avg_label": 0.4508064538240433,
|
| 1184 |
+
"epoch": 0.4548611111111111,
|
| 1185 |
+
"grad_norm": 0.8125,
|
| 1186 |
+
"learning_rate": 1.836421713086544e-05,
|
| 1187 |
+
"loss": 0.022061188239604235,
|
| 1188 |
+
"num_pairs": 28.0,
|
| 1189 |
+
"step": 131
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"avg_label": 0.4425466060638428,
|
| 1193 |
+
"epoch": 0.4583333333333333,
|
| 1194 |
+
"grad_norm": 1.5234375,
|
| 1195 |
+
"learning_rate": 1.83326010319797e-05,
|
| 1196 |
+
"loss": 0.08789217192679644,
|
| 1197 |
+
"num_pairs": 27.5,
|
| 1198 |
+
"step": 132
|
| 1199 |
+
},
|
| 1200 |
+
{
|
| 1201 |
+
"avg_label": 0.4583333283662796,
|
| 1202 |
+
"epoch": 0.4618055555555556,
|
| 1203 |
+
"grad_norm": 0.5859375,
|
| 1204 |
+
"learning_rate": 1.830071007755901e-05,
|
| 1205 |
+
"loss": 0.0123936936724931,
|
| 1206 |
+
"num_pairs": 28.0,
|
| 1207 |
+
"step": 133
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"avg_label": 0.4434974938631058,
|
| 1211 |
+
"epoch": 0.4652777777777778,
|
| 1212 |
+
"grad_norm": 0.640625,
|
| 1213 |
+
"learning_rate": 1.8268545319544443e-05,
|
| 1214 |
+
"loss": 0.0157788903452456,
|
| 1215 |
+
"num_pairs": 28.0,
|
| 1216 |
+
"step": 134
|
| 1217 |
+
},
|
| 1218 |
+
{
|
| 1219 |
+
"avg_label": 0.4349999725818634,
|
| 1220 |
+
"epoch": 0.46875,
|
| 1221 |
+
"grad_norm": 1.3046875,
|
| 1222 |
+
"learning_rate": 1.823610781890865e-05,
|
| 1223 |
+
"loss": 0.054201798513531685,
|
| 1224 |
+
"num_pairs": 26.0,
|
| 1225 |
+
"step": 135
|
| 1226 |
+
},
|
| 1227 |
+
{
|
| 1228 |
+
"avg_label": 0.4729166626930237,
|
| 1229 |
+
"epoch": 0.4722222222222222,
|
| 1230 |
+
"grad_norm": 0.875,
|
| 1231 |
+
"learning_rate": 1.820339864562085e-05,
|
| 1232 |
+
"loss": 0.03274301765486598,
|
| 1233 |
+
"num_pairs": 26.5,
|
| 1234 |
+
"step": 136
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"avg_label": 0.4567129611968994,
|
| 1238 |
+
"epoch": 0.4756944444444444,
|
| 1239 |
+
"grad_norm": 1.140625,
|
| 1240 |
+
"learning_rate": 1.817041887861153e-05,
|
| 1241 |
+
"loss": 0.05754423700273037,
|
| 1242 |
+
"num_pairs": 27.5,
|
| 1243 |
+
"step": 137
|
| 1244 |
+
},
|
| 1245 |
+
{
|
| 1246 |
+
"avg_label": 0.3948863595724106,
|
| 1247 |
+
"epoch": 0.4791666666666667,
|
| 1248 |
+
"grad_norm": 1.0703125,
|
| 1249 |
+
"learning_rate": 1.8137169605736867e-05,
|
| 1250 |
+
"loss": 0.07348897308111191,
|
| 1251 |
+
"num_pairs": 21.5,
|
| 1252 |
+
"step": 138
|
| 1253 |
+
},
|
| 1254 |
+
{
|
| 1255 |
+
"avg_label": 0.47046372294425964,
|
| 1256 |
+
"epoch": 0.4826388888888889,
|
| 1257 |
+
"grad_norm": 0.6015625,
|
| 1258 |
+
"learning_rate": 1.8103651923742846e-05,
|
| 1259 |
+
"loss": 0.04229694373498205,
|
| 1260 |
+
"num_pairs": 28.0,
|
| 1261 |
+
"step": 139
|
| 1262 |
+
},
|
| 1263 |
+
{
|
| 1264 |
+
"avg_label": 0.4568256735801697,
|
| 1265 |
+
"epoch": 0.4861111111111111,
|
| 1266 |
+
"grad_norm": 0.8046875,
|
| 1267 |
+
"learning_rate": 1.8069866938229066e-05,
|
| 1268 |
+
"loss": 0.0422638775780797,
|
| 1269 |
+
"num_pairs": 28.0,
|
| 1270 |
+
"step": 140
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"avg_label": 0.46851328015327454,
|
| 1274 |
+
"epoch": 0.4895833333333333,
|
| 1275 |
+
"grad_norm": 1.078125,
|
| 1276 |
+
"learning_rate": 1.8035815763612293e-05,
|
| 1277 |
+
"loss": 0.048252967186272144,
|
| 1278 |
+
"num_pairs": 27.5,
|
| 1279 |
+
"step": 141
|
| 1280 |
+
},
|
| 1281 |
+
{
|
| 1282 |
+
"avg_label": 0.46875,
|
| 1283 |
+
"epoch": 0.4930555555555556,
|
| 1284 |
+
"grad_norm": 1.109375,
|
| 1285 |
+
"learning_rate": 1.8001499523089683e-05,
|
| 1286 |
+
"loss": 0.022715996543411165,
|
| 1287 |
+
"num_pairs": 20.0,
|
| 1288 |
+
"step": 142
|
| 1289 |
+
},
|
| 1290 |
+
{
|
| 1291 |
+
"avg_label": 0.46064816415309906,
|
| 1292 |
+
"epoch": 0.4965277777777778,
|
| 1293 |
+
"grad_norm": 0.298828125,
|
| 1294 |
+
"learning_rate": 1.7966919348601754e-05,
|
| 1295 |
+
"loss": 0.0070006223395466805,
|
| 1296 |
+
"num_pairs": 28.0,
|
| 1297 |
+
"step": 143
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"avg_label": 0.4451013505458832,
|
| 1301 |
+
"epoch": 0.5,
|
| 1302 |
+
"grad_norm": 0.65625,
|
| 1303 |
+
"learning_rate": 1.7932076380795017e-05,
|
| 1304 |
+
"loss": 0.034362755715847015,
|
| 1305 |
+
"num_pairs": 28.0,
|
| 1306 |
+
"step": 144
|
| 1307 |
+
},
|
| 1308 |
+
{
|
| 1309 |
+
"avg_label": 0.4553571492433548,
|
| 1310 |
+
"epoch": 0.5034722222222222,
|
| 1311 |
+
"grad_norm": 0.6953125,
|
| 1312 |
+
"learning_rate": 1.7896971768984373e-05,
|
| 1313 |
+
"loss": 0.02424975112080574,
|
| 1314 |
+
"num_pairs": 28.0,
|
| 1315 |
+
"step": 145
|
| 1316 |
+
},
|
| 1317 |
+
{
|
| 1318 |
+
"avg_label": 0.40937501192092896,
|
| 1319 |
+
"epoch": 0.5069444444444444,
|
| 1320 |
+
"grad_norm": 1.40625,
|
| 1321 |
+
"learning_rate": 1.7861606671115207e-05,
|
| 1322 |
+
"loss": 0.09015845181420445,
|
| 1323 |
+
"num_pairs": 28.0,
|
| 1324 |
+
"step": 146
|
| 1325 |
+
},
|
| 1326 |
+
{
|
| 1327 |
+
"avg_label": 0.4572192281484604,
|
| 1328 |
+
"epoch": 0.5104166666666666,
|
| 1329 |
+
"grad_norm": 2.921875,
|
| 1330 |
+
"learning_rate": 1.7825982253725175e-05,
|
| 1331 |
+
"loss": 0.09106537594925612,
|
| 1332 |
+
"num_pairs": 28.0,
|
| 1333 |
+
"step": 147
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"avg_label": 0.4486243277788162,
|
| 1337 |
+
"epoch": 0.5138888888888888,
|
| 1338 |
+
"grad_norm": 0.22265625,
|
| 1339 |
+
"learning_rate": 1.7790099691905736e-05,
|
| 1340 |
+
"loss": 0.007117493078112602,
|
| 1341 |
+
"num_pairs": 28.0,
|
| 1342 |
+
"step": 148
|
| 1343 |
+
},
|
| 1344 |
+
{
|
| 1345 |
+
"avg_label": 0.42847222089767456,
|
| 1346 |
+
"epoch": 0.5173611111111112,
|
| 1347 |
+
"grad_norm": 0.326171875,
|
| 1348 |
+
"learning_rate": 1.7753960169263387e-05,
|
| 1349 |
+
"loss": 0.011324664112180471,
|
| 1350 |
+
"num_pairs": 25.5,
|
| 1351 |
+
"step": 149
|
| 1352 |
+
},
|
| 1353 |
+
{
|
| 1354 |
+
"avg_label": 0.439529225230217,
|
| 1355 |
+
"epoch": 0.5208333333333334,
|
| 1356 |
+
"grad_norm": 0.921875,
|
| 1357 |
+
"learning_rate": 1.7717564877880623e-05,
|
| 1358 |
+
"loss": 0.05978692602366209,
|
| 1359 |
+
"num_pairs": 27.5,
|
| 1360 |
+
"step": 150
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"avg_label": 0.4516952484846115,
|
| 1364 |
+
"epoch": 0.5243055555555556,
|
| 1365 |
+
"grad_norm": 1.375,
|
| 1366 |
+
"learning_rate": 1.7680915018276613e-05,
|
| 1367 |
+
"loss": 0.050734762102365494,
|
| 1368 |
+
"num_pairs": 28.0,
|
| 1369 |
+
"step": 151
|
| 1370 |
+
},
|
| 1371 |
+
{
|
| 1372 |
+
"avg_label": 0.41067710518836975,
|
| 1373 |
+
"epoch": 0.5277777777777778,
|
| 1374 |
+
"grad_norm": 0.93359375,
|
| 1375 |
+
"learning_rate": 1.764401179936761e-05,
|
| 1376 |
+
"loss": 0.0304207439039601,
|
| 1377 |
+
"num_pairs": 27.5,
|
| 1378 |
+
"step": 152
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"avg_label": 0.491695836186409,
|
| 1382 |
+
"epoch": 0.53125,
|
| 1383 |
+
"grad_norm": 0.482421875,
|
| 1384 |
+
"learning_rate": 1.7606856438427054e-05,
|
| 1385 |
+
"loss": 0.014743164298124611,
|
| 1386 |
+
"num_pairs": 27.5,
|
| 1387 |
+
"step": 153
|
| 1388 |
+
},
|
| 1389 |
+
{
|
| 1390 |
+
"avg_label": 0.446428582072258,
|
| 1391 |
+
"epoch": 0.5347222222222222,
|
| 1392 |
+
"grad_norm": 0.890625,
|
| 1393 |
+
"learning_rate": 1.7569450161045444e-05,
|
| 1394 |
+
"loss": 0.04068564437329769,
|
| 1395 |
+
"num_pairs": 23.5,
|
| 1396 |
+
"step": 154
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"avg_label": 0.45932111144065857,
|
| 1400 |
+
"epoch": 0.5381944444444444,
|
| 1401 |
+
"grad_norm": 0.443359375,
|
| 1402 |
+
"learning_rate": 1.7531794201089888e-05,
|
| 1403 |
+
"loss": 0.022642125841230154,
|
| 1404 |
+
"num_pairs": 28.0,
|
| 1405 |
+
"step": 155
|
| 1406 |
+
},
|
| 1407 |
+
{
|
| 1408 |
+
"avg_label": 0.45102816820144653,
|
| 1409 |
+
"epoch": 0.5416666666666666,
|
| 1410 |
+
"grad_norm": 0.333984375,
|
| 1411 |
+
"learning_rate": 1.749388980066342e-05,
|
| 1412 |
+
"loss": 0.009190058917738497,
|
| 1413 |
+
"num_pairs": 28.0,
|
| 1414 |
+
"step": 156
|
| 1415 |
+
},
|
| 1416 |
+
{
|
| 1417 |
+
"avg_label": 0.44252873957157135,
|
| 1418 |
+
"epoch": 0.5451388888888888,
|
| 1419 |
+
"grad_norm": 0.490234375,
|
| 1420 |
+
"learning_rate": 1.745573821006403e-05,
|
| 1421 |
+
"loss": 0.0138811687938869,
|
| 1422 |
+
"num_pairs": 27.5,
|
| 1423 |
+
"step": 157
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"avg_label": 0.44221231341362,
|
| 1427 |
+
"epoch": 0.5486111111111112,
|
| 1428 |
+
"grad_norm": 1.2890625,
|
| 1429 |
+
"learning_rate": 1.7417340687743393e-05,
|
| 1430 |
+
"loss": 0.05438859760761261,
|
| 1431 |
+
"num_pairs": 28.0,
|
| 1432 |
+
"step": 158
|
| 1433 |
+
},
|
| 1434 |
+
{
|
| 1435 |
+
"avg_label": 0.4568452388048172,
|
| 1436 |
+
"epoch": 0.5520833333333334,
|
| 1437 |
+
"grad_norm": 0.25,
|
| 1438 |
+
"learning_rate": 1.7378698500265402e-05,
|
| 1439 |
+
"loss": 0.010841536102816463,
|
| 1440 |
+
"num_pairs": 28.0,
|
| 1441 |
+
"step": 159
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"avg_label": 0.4703125059604645,
|
| 1445 |
+
"epoch": 0.5555555555555556,
|
| 1446 |
+
"grad_norm": 0.8515625,
|
| 1447 |
+
"learning_rate": 1.7339812922264366e-05,
|
| 1448 |
+
"loss": 0.03851235564798117,
|
| 1449 |
+
"num_pairs": 28.0,
|
| 1450 |
+
"step": 160
|
| 1451 |
+
},
|
| 1452 |
+
{
|
| 1453 |
+
"avg_label": 0.45864899456501007,
|
| 1454 |
+
"epoch": 0.5590277777777778,
|
| 1455 |
+
"grad_norm": 3.4375,
|
| 1456 |
+
"learning_rate": 1.730068523640295e-05,
|
| 1457 |
+
"loss": 0.11844043270684779,
|
| 1458 |
+
"num_pairs": 26.5,
|
| 1459 |
+
"step": 161
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"avg_label": 0.4791666567325592,
|
| 1463 |
+
"epoch": 0.5625,
|
| 1464 |
+
"grad_norm": 0.224609375,
|
| 1465 |
+
"learning_rate": 1.72613167333299e-05,
|
| 1466 |
+
"loss": 0.028862916253274307,
|
| 1467 |
+
"num_pairs": 24.0,
|
| 1468 |
+
"step": 162
|
| 1469 |
+
},
|
| 1470 |
+
{
|
| 1471 |
+
"avg_label": 0.45817309617996216,
|
| 1472 |
+
"epoch": 0.5659722222222222,
|
| 1473 |
+
"grad_norm": 0.2734375,
|
| 1474 |
+
"learning_rate": 1.7221708711637455e-05,
|
| 1475 |
+
"loss": 0.00937138102017343,
|
| 1476 |
+
"num_pairs": 27.5,
|
| 1477 |
+
"step": 163
|
| 1478 |
+
},
|
| 1479 |
+
{
|
| 1480 |
+
"avg_label": 0.43892045319080353,
|
| 1481 |
+
"epoch": 0.5694444444444444,
|
| 1482 |
+
"grad_norm": 0.671875,
|
| 1483 |
+
"learning_rate": 1.718186247781849e-05,
|
| 1484 |
+
"loss": 0.028468840289860964,
|
| 1485 |
+
"num_pairs": 28.0,
|
| 1486 |
+
"step": 164
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"avg_label": 0.4518229067325592,
|
| 1490 |
+
"epoch": 0.5729166666666666,
|
| 1491 |
+
"grad_norm": 6.125,
|
| 1492 |
+
"learning_rate": 1.7141779346223465e-05,
|
| 1493 |
+
"loss": 0.30780595541000366,
|
| 1494 |
+
"num_pairs": 27.5,
|
| 1495 |
+
"step": 165
|
| 1496 |
+
},
|
| 1497 |
+
{
|
| 1498 |
+
"avg_label": 0.45110294222831726,
|
| 1499 |
+
"epoch": 0.5763888888888888,
|
| 1500 |
+
"grad_norm": 0.306640625,
|
| 1501 |
+
"learning_rate": 1.7101460639017034e-05,
|
| 1502 |
+
"loss": 0.008247810648754239,
|
| 1503 |
+
"num_pairs": 28.0,
|
| 1504 |
+
"step": 166
|
| 1505 |
+
},
|
| 1506 |
+
{
|
| 1507 |
+
"avg_label": 0.46600276231765747,
|
| 1508 |
+
"epoch": 0.5798611111111112,
|
| 1509 |
+
"grad_norm": 0.8125,
|
| 1510 |
+
"learning_rate": 1.7060907686134445e-05,
|
| 1511 |
+
"loss": 0.06383802928030491,
|
| 1512 |
+
"num_pairs": 28.0,
|
| 1513 |
+
"step": 167
|
| 1514 |
+
},
|
| 1515 |
+
{
|
| 1516 |
+
"avg_label": 0.4557291716337204,
|
| 1517 |
+
"epoch": 0.5833333333333334,
|
| 1518 |
+
"grad_norm": 0.40234375,
|
| 1519 |
+
"learning_rate": 1.7020121825237672e-05,
|
| 1520 |
+
"loss": 0.018603924312628806,
|
| 1521 |
+
"num_pairs": 25.5,
|
| 1522 |
+
"step": 168
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"avg_label": 0.4603828638792038,
|
| 1526 |
+
"epoch": 0.5868055555555556,
|
| 1527 |
+
"grad_norm": 0.82421875,
|
| 1528 |
+
"learning_rate": 1.6979104401671296e-05,
|
| 1529 |
+
"loss": 0.022878690622746944,
|
| 1530 |
+
"num_pairs": 28.0,
|
| 1531 |
+
"step": 169
|
| 1532 |
+
},
|
| 1533 |
+
{
|
| 1534 |
+
"avg_label": 0.4588768184185028,
|
| 1535 |
+
"epoch": 0.5902777777777778,
|
| 1536 |
+
"grad_norm": 1.4375,
|
| 1537 |
+
"learning_rate": 1.693785676841812e-05,
|
| 1538 |
+
"loss": 0.06640730798244476,
|
| 1539 |
+
"num_pairs": 28.0,
|
| 1540 |
+
"step": 170
|
| 1541 |
+
},
|
| 1542 |
+
{
|
| 1543 |
+
"avg_label": 0.4221428632736206,
|
| 1544 |
+
"epoch": 0.59375,
|
| 1545 |
+
"grad_norm": 0.107421875,
|
| 1546 |
+
"learning_rate": 1.6896380286054537e-05,
|
| 1547 |
+
"loss": 0.0038028154522180557,
|
| 1548 |
+
"num_pairs": 28.0,
|
| 1549 |
+
"step": 171
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"avg_label": 0.47298388183116913,
|
| 1553 |
+
"epoch": 0.5972222222222222,
|
| 1554 |
+
"grad_norm": 0.55078125,
|
| 1555 |
+
"learning_rate": 1.6854676322705673e-05,
|
| 1556 |
+
"loss": 0.019407693296670914,
|
| 1557 |
+
"num_pairs": 28.0,
|
| 1558 |
+
"step": 172
|
| 1559 |
+
},
|
| 1560 |
+
{
|
| 1561 |
+
"avg_label": 0.444306343793869,
|
| 1562 |
+
"epoch": 0.6006944444444444,
|
| 1563 |
+
"grad_norm": 2.65625,
|
| 1564 |
+
"learning_rate": 1.6812746254000222e-05,
|
| 1565 |
+
"loss": 0.11606021597981453,
|
| 1566 |
+
"num_pairs": 28.0,
|
| 1567 |
+
"step": 173
|
| 1568 |
+
},
|
| 1569 |
+
{
|
| 1570 |
+
"avg_label": 0.3993750214576721,
|
| 1571 |
+
"epoch": 0.6041666666666666,
|
| 1572 |
+
"grad_norm": 1.921875,
|
| 1573 |
+
"learning_rate": 1.67705914630251e-05,
|
| 1574 |
+
"loss": 0.08407180290669203,
|
| 1575 |
+
"num_pairs": 27.0,
|
| 1576 |
+
"step": 174
|
| 1577 |
+
},
|
| 1578 |
+
{
|
| 1579 |
+
"avg_label": 0.4376724064350128,
|
| 1580 |
+
"epoch": 0.6076388888888888,
|
| 1581 |
+
"grad_norm": 0.59765625,
|
| 1582 |
+
"learning_rate": 1.6728213340279822e-05,
|
| 1583 |
+
"loss": 0.01655407203361392,
|
| 1584 |
+
"num_pairs": 28.0,
|
| 1585 |
+
"step": 175
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"avg_label": 0.31534090638160706,
|
| 1589 |
+
"epoch": 0.6111111111111112,
|
| 1590 |
+
"grad_norm": 0.1396484375,
|
| 1591 |
+
"learning_rate": 1.668561328363061e-05,
|
| 1592 |
+
"loss": 0.007483100518584251,
|
| 1593 |
+
"num_pairs": 20.0,
|
| 1594 |
+
"step": 176
|
| 1595 |
+
},
|
| 1596 |
+
{
|
| 1597 |
+
"avg_label": 0.4442349076271057,
|
| 1598 |
+
"epoch": 0.6145833333333334,
|
| 1599 |
+
"grad_norm": 0.7265625,
|
| 1600 |
+
"learning_rate": 1.6642792698264313e-05,
|
| 1601 |
+
"loss": 0.03157099802047014,
|
| 1602 |
+
"num_pairs": 28.0,
|
| 1603 |
+
"step": 177
|
| 1604 |
+
},
|
| 1605 |
+
{
|
| 1606 |
+
"avg_label": 0.4552556872367859,
|
| 1607 |
+
"epoch": 0.6180555555555556,
|
| 1608 |
+
"grad_norm": 1.0234375,
|
| 1609 |
+
"learning_rate": 1.6599752996642044e-05,
|
| 1610 |
+
"loss": 0.0451383707113564,
|
| 1611 |
+
"num_pairs": 28.0,
|
| 1612 |
+
"step": 178
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"avg_label": 0.4735576957464218,
|
| 1616 |
+
"epoch": 0.6215277777777778,
|
| 1617 |
+
"grad_norm": 1.203125,
|
| 1618 |
+
"learning_rate": 1.655649559845258e-05,
|
| 1619 |
+
"loss": 0.056825135834515095,
|
| 1620 |
+
"num_pairs": 22.0,
|
| 1621 |
+
"step": 179
|
| 1622 |
+
},
|
| 1623 |
+
{
|
| 1624 |
+
"avg_label": 0.38331981003284454,
|
| 1625 |
+
"epoch": 0.625,
|
| 1626 |
+
"grad_norm": 0.59375,
|
| 1627 |
+
"learning_rate": 1.651302193056555e-05,
|
| 1628 |
+
"loss": 0.03129632119089365,
|
| 1629 |
+
"num_pairs": 22.5,
|
| 1630 |
+
"step": 180
|
| 1631 |
+
},
|
| 1632 |
+
{
|
| 1633 |
+
"avg_label": 0.4483432173728943,
|
| 1634 |
+
"epoch": 0.6284722222222222,
|
| 1635 |
+
"grad_norm": 1.34375,
|
| 1636 |
+
"learning_rate": 1.6469333426984357e-05,
|
| 1637 |
+
"loss": 0.05972203053534031,
|
| 1638 |
+
"num_pairs": 28.0,
|
| 1639 |
+
"step": 181
|
| 1640 |
+
},
|
| 1641 |
+
{
|
| 1642 |
+
"avg_label": 0.468173086643219,
|
| 1643 |
+
"epoch": 0.6319444444444444,
|
| 1644 |
+
"grad_norm": 1.4765625,
|
| 1645 |
+
"learning_rate": 1.6425431528798883e-05,
|
| 1646 |
+
"loss": 0.06996654625982046,
|
| 1647 |
+
"num_pairs": 28.0,
|
| 1648 |
+
"step": 182
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"avg_label": 0.4802695959806442,
|
| 1652 |
+
"epoch": 0.6354166666666666,
|
| 1653 |
+
"grad_norm": 0.890625,
|
| 1654 |
+
"learning_rate": 1.6381317684137946e-05,
|
| 1655 |
+
"loss": 0.038833396742120385,
|
| 1656 |
+
"num_pairs": 27.5,
|
| 1657 |
+
"step": 183
|
| 1658 |
+
},
|
| 1659 |
+
{
|
| 1660 |
+
"avg_label": 0.44396552443504333,
|
| 1661 |
+
"epoch": 0.6388888888888888,
|
| 1662 |
+
"grad_norm": 0.33984375,
|
| 1663 |
+
"learning_rate": 1.6336993348121543e-05,
|
| 1664 |
+
"loss": 0.010088181821629405,
|
| 1665 |
+
"num_pairs": 27.5,
|
| 1666 |
+
"step": 184
|
| 1667 |
+
},
|
| 1668 |
+
{
|
| 1669 |
+
"avg_label": 0.45236895978450775,
|
| 1670 |
+
"epoch": 0.6423611111111112,
|
| 1671 |
+
"grad_norm": 0.10546875,
|
| 1672 |
+
"learning_rate": 1.6292459982812845e-05,
|
| 1673 |
+
"loss": 0.0034077019081451,
|
| 1674 |
+
"num_pairs": 28.0,
|
| 1675 |
+
"step": 185
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"avg_label": 0.47602301836013794,
|
| 1679 |
+
"epoch": 0.6458333333333334,
|
| 1680 |
+
"grad_norm": 0.25,
|
| 1681 |
+
"learning_rate": 1.624771905716997e-05,
|
| 1682 |
+
"loss": 0.004801646777195856,
|
| 1683 |
+
"num_pairs": 27.0,
|
| 1684 |
+
"step": 186
|
| 1685 |
+
},
|
| 1686 |
+
{
|
| 1687 |
+
"avg_label": 0.45323528349399567,
|
| 1688 |
+
"epoch": 0.6493055555555556,
|
| 1689 |
+
"grad_norm": 0.59375,
|
| 1690 |
+
"learning_rate": 1.620277204699754e-05,
|
| 1691 |
+
"loss": 0.042771865613758564,
|
| 1692 |
+
"num_pairs": 27.5,
|
| 1693 |
+
"step": 187
|
| 1694 |
+
},
|
| 1695 |
+
{
|
| 1696 |
+
"avg_label": 0.4444444477558136,
|
| 1697 |
+
"epoch": 0.6527777777777778,
|
| 1698 |
+
"grad_norm": 0.6640625,
|
| 1699 |
+
"learning_rate": 1.615762043489797e-05,
|
| 1700 |
+
"loss": 0.0236468983348459,
|
| 1701 |
+
"num_pairs": 25.0,
|
| 1702 |
+
"step": 188
|
| 1703 |
+
},
|
| 1704 |
+
{
|
| 1705 |
+
"avg_label": 0.44826723635196686,
|
| 1706 |
+
"epoch": 0.65625,
|
| 1707 |
+
"grad_norm": 0.1982421875,
|
| 1708 |
+
"learning_rate": 1.611226571022261e-05,
|
| 1709 |
+
"loss": 0.007521128980442882,
|
| 1710 |
+
"num_pairs": 28.0,
|
| 1711 |
+
"step": 189
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"avg_label": 0.4715389013290405,
|
| 1715 |
+
"epoch": 0.6597222222222222,
|
| 1716 |
+
"grad_norm": 0.5625,
|
| 1717 |
+
"learning_rate": 1.6066709369022576e-05,
|
| 1718 |
+
"loss": 0.020246338099241257,
|
| 1719 |
+
"num_pairs": 27.0,
|
| 1720 |
+
"step": 190
|
| 1721 |
+
},
|
| 1722 |
+
{
|
| 1723 |
+
"avg_label": 0.46243055164813995,
|
| 1724 |
+
"epoch": 0.6631944444444444,
|
| 1725 |
+
"grad_norm": 3.53125,
|
| 1726 |
+
"learning_rate": 1.6020952913999423e-05,
|
| 1727 |
+
"loss": 0.08208651561290026,
|
| 1728 |
+
"num_pairs": 27.5,
|
| 1729 |
+
"step": 191
|
| 1730 |
+
},
|
| 1731 |
+
{
|
| 1732 |
+
"avg_label": 0.4524739533662796,
|
| 1733 |
+
"epoch": 0.6666666666666666,
|
| 1734 |
+
"grad_norm": 0.166015625,
|
| 1735 |
+
"learning_rate": 1.5974997854455575e-05,
|
| 1736 |
+
"loss": 0.0023918856459204108,
|
| 1737 |
+
"num_pairs": 28.0,
|
| 1738 |
+
"step": 192
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"avg_label": 0.46304087340831757,
|
| 1742 |
+
"epoch": 0.6701388888888888,
|
| 1743 |
+
"grad_norm": 0.412109375,
|
| 1744 |
+
"learning_rate": 1.5928845706244537e-05,
|
| 1745 |
+
"loss": 0.020947942975908518,
|
| 1746 |
+
"num_pairs": 28.0,
|
| 1747 |
+
"step": 193
|
| 1748 |
+
},
|
| 1749 |
+
{
|
| 1750 |
+
"avg_label": 0.4618534445762634,
|
| 1751 |
+
"epoch": 0.6736111111111112,
|
| 1752 |
+
"grad_norm": 0.232421875,
|
| 1753 |
+
"learning_rate": 1.588249799172089e-05,
|
| 1754 |
+
"loss": 0.005286574247293174,
|
| 1755 |
+
"num_pairs": 27.5,
|
| 1756 |
+
"step": 194
|
| 1757 |
+
},
|
| 1758 |
+
{
|
| 1759 |
+
"avg_label": 0.4678194969892502,
|
| 1760 |
+
"epoch": 0.6770833333333334,
|
| 1761 |
+
"grad_norm": 1.1015625,
|
| 1762 |
+
"learning_rate": 1.583595623969009e-05,
|
| 1763 |
+
"loss": 0.05471465736627579,
|
| 1764 |
+
"num_pairs": 28.0,
|
| 1765 |
+
"step": 195
|
| 1766 |
+
},
|
| 1767 |
+
{
|
| 1768 |
+
"avg_label": 0.4570639431476593,
|
| 1769 |
+
"epoch": 0.6805555555555556,
|
| 1770 |
+
"grad_norm": 0.2197265625,
|
| 1771 |
+
"learning_rate": 1.5789221985358017e-05,
|
| 1772 |
+
"loss": 0.004043272638227791,
|
| 1773 |
+
"num_pairs": 27.5,
|
| 1774 |
+
"step": 196
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"avg_label": 0.44032806158065796,
|
| 1778 |
+
"epoch": 0.6840277777777778,
|
| 1779 |
+
"grad_norm": 2.171875,
|
| 1780 |
+
"learning_rate": 1.574229677028036e-05,
|
| 1781 |
+
"loss": 0.03309955738950521,
|
| 1782 |
+
"num_pairs": 25.5,
|
| 1783 |
+
"step": 197
|
| 1784 |
+
},
|
| 1785 |
+
{
|
| 1786 |
+
"avg_label": 0.4401831030845642,
|
| 1787 |
+
"epoch": 0.6875,
|
| 1788 |
+
"grad_norm": 1.390625,
|
| 1789 |
+
"learning_rate": 1.5695182142311743e-05,
|
| 1790 |
+
"loss": 0.03160541597753763,
|
| 1791 |
+
"num_pairs": 28.0,
|
| 1792 |
+
"step": 198
|
| 1793 |
+
},
|
| 1794 |
+
{
|
| 1795 |
+
"avg_label": 0.45328740775585175,
|
| 1796 |
+
"epoch": 0.6909722222222222,
|
| 1797 |
+
"grad_norm": 4.3125,
|
| 1798 |
+
"learning_rate": 1.564787965555469e-05,
|
| 1799 |
+
"loss": 0.07159067690372467,
|
| 1800 |
+
"num_pairs": 27.5,
|
| 1801 |
+
"step": 199
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"avg_label": 0.4379340261220932,
|
| 1805 |
+
"epoch": 0.6944444444444444,
|
| 1806 |
+
"grad_norm": 0.65625,
|
| 1807 |
+
"learning_rate": 1.560039087030836e-05,
|
| 1808 |
+
"loss": 0.0306890313513577,
|
| 1809 |
+
"num_pairs": 28.0,
|
| 1810 |
+
"step": 200
|
| 1811 |
+
}
|
| 1812 |
+
],
|
| 1813 |
+
"logging_steps": 1.0,
|
| 1814 |
+
"max_steps": 576,
|
| 1815 |
+
"num_input_tokens_seen": 0,
|
| 1816 |
+
"num_train_epochs": 2,
|
| 1817 |
+
"save_steps": 100,
|
| 1818 |
+
"stateful_callbacks": {
|
| 1819 |
+
"TrainerControl": {
|
| 1820 |
+
"args": {
|
| 1821 |
+
"should_epoch_stop": false,
|
| 1822 |
+
"should_evaluate": false,
|
| 1823 |
+
"should_log": false,
|
| 1824 |
+
"should_save": true,
|
| 1825 |
+
"should_training_stop": false
|
| 1826 |
+
},
|
| 1827 |
+
"attributes": {}
|
| 1828 |
+
}
|
| 1829 |
+
},
|
| 1830 |
+
"total_flos": 0.0,
|
| 1831 |
+
"train_batch_size": 32,
|
| 1832 |
+
"trial_name": null,
|
| 1833 |
+
"trial_params": null
|
| 1834 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ec0a2c7e7e550f7fac2bfb9f535d02f16d9428695d17f831a7aaf9a829592c3f
|
| 3 |
+
size 5496
|
vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|