test-5-checkpoint-500
Browse files- test-5-checkpoint-500/config.json +187 -0
- test-5-checkpoint-500/dataset_statistics.json +253 -0
- test-5-checkpoint-500/model.safetensors +3 -0
- test-5-checkpoint-500/optimizer.pt +3 -0
- test-5-checkpoint-500/rng_state_0.pth +3 -0
- test-5-checkpoint-500/rng_state_1.pth +3 -0
- test-5-checkpoint-500/rng_state_2.pth +3 -0
- test-5-checkpoint-500/rng_state_3.pth +3 -0
- test-5-checkpoint-500/rng_state_4.pth +3 -0
- test-5-checkpoint-500/rng_state_5.pth +3 -0
- test-5-checkpoint-500/rng_state_6.pth +3 -0
- test-5-checkpoint-500/rng_state_7.pth +3 -0
- test-5-checkpoint-500/scheduler.pt +3 -0
- test-5-checkpoint-500/trainer_state.json +108 -0
- test-5-checkpoint-500/training_args.bin +3 -0
test-5-checkpoint-500/config.json
ADDED
|
@@ -0,0 +1,187 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"batch_size": 64,
|
| 3 |
+
"buffer_size": 64,
|
| 4 |
+
"eval_mix": "chris_aubo",
|
| 5 |
+
"frozen_keys": [
|
| 6 |
+
"*hf_model*"
|
| 7 |
+
],
|
| 8 |
+
"lora": false,
|
| 9 |
+
"lora_config": {
|
| 10 |
+
"bias": "none",
|
| 11 |
+
"lora_alpha": 16,
|
| 12 |
+
"lora_dropout": 0.05,
|
| 13 |
+
"r": 8,
|
| 14 |
+
"target_modules": [
|
| 15 |
+
"module.octo_transformer.transformer.transformer.encoder_layers.0.attention.out_proj",
|
| 16 |
+
"module.octo_transformer.transformer.transformer.encoder_layers.0.mlp_block.mlp.0",
|
| 17 |
+
"module.octo_transformer.transformer.transformer.encoder_layers.0.mlp_block.mlp.3",
|
| 18 |
+
"module.octo_transformer.transformer.transformer.encoder_layers.1.attention.out_proj",
|
| 19 |
+
"module.octo_transformer.transformer.transformer.encoder_layers.1.mlp_block.mlp.0",
|
| 20 |
+
"module.octo_transformer.transformer.transformer.encoder_layers.1.mlp_block.mlp.3",
|
| 21 |
+
"module.octo_transformer.transformer.transformer.encoder_layers.2.attention.out_proj",
|
| 22 |
+
"module.octo_transformer.transformer.transformer.encoder_layers.2.mlp_block.mlp.0",
|
| 23 |
+
"module.octo_transformer.transformer.transformer.encoder_layers.2.mlp_block.mlp.3",
|
| 24 |
+
"module.octo_transformer.transformer.transformer.encoder_layers.3.attention.out_proj",
|
| 25 |
+
"module.octo_transformer.transformer.transformer.encoder_layers.3.mlp_block.mlp.0",
|
| 26 |
+
"module.octo_transformer.transformer.transformer.encoder_layers.3.mlp_block.mlp.3",
|
| 27 |
+
"module.octo_transformer.transformer.transformer.encoder_layers.4.attention.out_proj",
|
| 28 |
+
"module.octo_transformer.transformer.transformer.encoder_layers.4.mlp_block.mlp.0",
|
| 29 |
+
"module.octo_transformer.transformer.transformer.encoder_layers.4.mlp_block.mlp.3",
|
| 30 |
+
"module.octo_transformer.transformer.transformer.encoder_layers.5.attention.out_proj",
|
| 31 |
+
"module.octo_transformer.transformer.transformer.encoder_layers.5.mlp_block.mlp.0",
|
| 32 |
+
"module.octo_transformer.transformer.transformer.encoder_layers.5.mlp_block.mlp.3",
|
| 33 |
+
"module.octo_transformer.transformer.transformer.encoder_layers.6.attention.out_proj",
|
| 34 |
+
"module.octo_transformer.transformer.transformer.encoder_layers.6.mlp_block.mlp.0",
|
| 35 |
+
"module.octo_transformer.transformer.transformer.encoder_layers.6.mlp_block.mlp.3",
|
| 36 |
+
"module.octo_transformer.transformer.transformer.encoder_layers.7.attention.out_proj",
|
| 37 |
+
"module.octo_transformer.transformer.transformer.encoder_layers.7.mlp_block.mlp.0",
|
| 38 |
+
"module.octo_transformer.transformer.transformer.encoder_layers.7.mlp_block.mlp.3",
|
| 39 |
+
"module.octo_transformer.transformer.transformer.encoder_layers.8.attention.out_proj",
|
| 40 |
+
"module.octo_transformer.transformer.transformer.encoder_layers.8.mlp_block.mlp.0",
|
| 41 |
+
"module.octo_transformer.transformer.transformer.encoder_layers.8.mlp_block.mlp.3",
|
| 42 |
+
"module.octo_transformer.transformer.transformer.encoder_layers.9.attention.out_proj",
|
| 43 |
+
"module.octo_transformer.transformer.transformer.encoder_layers.9.mlp_block.mlp.0",
|
| 44 |
+
"module.octo_transformer.transformer.transformer.encoder_layers.9.mlp_block.mlp.3",
|
| 45 |
+
"module.octo_transformer.transformer.transformer.encoder_layers.10.attention.out_proj",
|
| 46 |
+
"module.octo_transformer.transformer.transformer.encoder_layers.10.mlp_block.mlp.0",
|
| 47 |
+
"module.octo_transformer.transformer.transformer.encoder_layers.10.mlp_block.mlp.3",
|
| 48 |
+
"module.octo_transformer.transformer.transformer.encoder_layers.11.attention.out_proj",
|
| 49 |
+
"module.octo_transformer.transformer.transformer.encoder_layers.11.mlp_block.mlp.0",
|
| 50 |
+
"module.octo_transformer.transformer.transformer.encoder_layers.11.mlp_block.mlp.3"
|
| 51 |
+
]
|
| 52 |
+
},
|
| 53 |
+
"model": {
|
| 54 |
+
"heads": {
|
| 55 |
+
"action": {
|
| 56 |
+
"args": [],
|
| 57 |
+
"kwargs": {
|
| 58 |
+
"action_dim": 7,
|
| 59 |
+
"action_horizon": 4,
|
| 60 |
+
"dropout_rate": 0.0,
|
| 61 |
+
"n_diffusion_samples": 1,
|
| 62 |
+
"readout_key": "readout_action",
|
| 63 |
+
"token_embedding_size": 768,
|
| 64 |
+
"use_map": false
|
| 65 |
+
},
|
| 66 |
+
"module": "octo.model.components.action_heads",
|
| 67 |
+
"name": "DiffusionActionHead"
|
| 68 |
+
}
|
| 69 |
+
},
|
| 70 |
+
"max_horizon": 10,
|
| 71 |
+
"observation_tokenizers": {
|
| 72 |
+
"primary": {
|
| 73 |
+
"args": [],
|
| 74 |
+
"kwargs": {
|
| 75 |
+
"encoder": {
|
| 76 |
+
"args": [],
|
| 77 |
+
"kwargs": {
|
| 78 |
+
"in_features": 6
|
| 79 |
+
},
|
| 80 |
+
"module": "octo.model.components.vit_encoders",
|
| 81 |
+
"name": "SmallStem16"
|
| 82 |
+
},
|
| 83 |
+
"obs_stack_keys": [
|
| 84 |
+
"image_primary"
|
| 85 |
+
],
|
| 86 |
+
"task_stack_keys": [
|
| 87 |
+
"image_primary"
|
| 88 |
+
]
|
| 89 |
+
},
|
| 90 |
+
"module": "octo.model.components.tokenizers",
|
| 91 |
+
"name": "ImageTokenizer"
|
| 92 |
+
}
|
| 93 |
+
},
|
| 94 |
+
"readouts": {
|
| 95 |
+
"action": 1
|
| 96 |
+
},
|
| 97 |
+
"repeat_task_tokens": true,
|
| 98 |
+
"task_tokenizers": {
|
| 99 |
+
"language": {
|
| 100 |
+
"args": [],
|
| 101 |
+
"kwargs": {
|
| 102 |
+
"encoder": "t5-base",
|
| 103 |
+
"finetune_encoder": false
|
| 104 |
+
},
|
| 105 |
+
"module": "octo.model.components.tokenizers",
|
| 106 |
+
"name": "LanguageTokenizer"
|
| 107 |
+
}
|
| 108 |
+
},
|
| 109 |
+
"token_embedding_size": 768,
|
| 110 |
+
"transformer_kwargs": {
|
| 111 |
+
"add_position_embedding": false,
|
| 112 |
+
"attention_dropout_rate": 0.0,
|
| 113 |
+
"dropout_rate": 0.0,
|
| 114 |
+
"mlp_dim": 3072,
|
| 115 |
+
"num_attention_heads": 12,
|
| 116 |
+
"num_layers": 12
|
| 117 |
+
},
|
| 118 |
+
"use_correct_attention": true
|
| 119 |
+
},
|
| 120 |
+
"normalize_method": "sign",
|
| 121 |
+
"obs_token_nums": {
|
| 122 |
+
"primary": 256
|
| 123 |
+
},
|
| 124 |
+
"output_dir": "octo_base_aubo_checkpoints",
|
| 125 |
+
"pretrained": {
|
| 126 |
+
"kwargs": {
|
| 127 |
+
"subpath": "oxe-g2-checkpoint-300000"
|
| 128 |
+
},
|
| 129 |
+
"model": "hf://chuanmew/octo_torch"
|
| 130 |
+
},
|
| 131 |
+
"run_name": "aubo",
|
| 132 |
+
"sampler_num_samples": 2560000,
|
| 133 |
+
"seed": 2,
|
| 134 |
+
"subsample_length": 99999,
|
| 135 |
+
"text_processor": {
|
| 136 |
+
"args": [],
|
| 137 |
+
"kwargs": {
|
| 138 |
+
"encode_with_model": false,
|
| 139 |
+
"tokenizer_kwargs": {
|
| 140 |
+
"max_length": 16,
|
| 141 |
+
"padding": "max_length",
|
| 142 |
+
"return_tensors": "np",
|
| 143 |
+
"truncation": true
|
| 144 |
+
},
|
| 145 |
+
"tokenizer_name": "t5-base"
|
| 146 |
+
},
|
| 147 |
+
"module": "octo.components.text_processing",
|
| 148 |
+
"name": "HFTokenizer"
|
| 149 |
+
},
|
| 150 |
+
"train_mix": "czy_aubo",
|
| 151 |
+
"training_arguments": {
|
| 152 |
+
"bf16": true,
|
| 153 |
+
"dataloader_pin_memory": true,
|
| 154 |
+
"ddp_find_unused_parameters": true,
|
| 155 |
+
"eval_steps": 100,
|
| 156 |
+
"gradient_accumulation_steps": 4,
|
| 157 |
+
"learning_rate": 5e-05,
|
| 158 |
+
"logging_nan_inf_filter": false,
|
| 159 |
+
"logging_steps": 100,
|
| 160 |
+
"lr_scheduler_kwargs": {
|
| 161 |
+
"decay_type": "cosine",
|
| 162 |
+
"min_lr_ratio": 0.05,
|
| 163 |
+
"num_decay_steps": 298000
|
| 164 |
+
},
|
| 165 |
+
"lr_scheduler_type": "constant",
|
| 166 |
+
"max_grad_norm": 1.0,
|
| 167 |
+
"max_steps": 300000,
|
| 168 |
+
"optim": "adamw_torch_fused",
|
| 169 |
+
"per_device_eval_batch_size": 1,
|
| 170 |
+
"per_device_train_batch_size": 1,
|
| 171 |
+
"report_to": "wandb",
|
| 172 |
+
"save_safetensors": true,
|
| 173 |
+
"save_steps": 500,
|
| 174 |
+
"torch_compile": true,
|
| 175 |
+
"warmup_steps": 2000,
|
| 176 |
+
"weight_decay": 0.1
|
| 177 |
+
},
|
| 178 |
+
"training_keys": [
|
| 179 |
+
"*action*",
|
| 180 |
+
"*task_projections*",
|
| 181 |
+
"*obs_projections*",
|
| 182 |
+
"*readout_embeddings*",
|
| 183 |
+
"*task_pos_embeddings*",
|
| 184 |
+
"*obs_pos_embeddings*",
|
| 185 |
+
"*observation_tokenizers*"
|
| 186 |
+
]
|
| 187 |
+
}
|
test-5-checkpoint-500/dataset_statistics.json
ADDED
|
@@ -0,0 +1,253 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"task_index": {
|
| 4 |
+
"min": [
|
| 5 |
+
0
|
| 6 |
+
],
|
| 7 |
+
"max": [
|
| 8 |
+
6
|
| 9 |
+
],
|
| 10 |
+
"mean": [
|
| 11 |
+
2.837244511733535
|
| 12 |
+
],
|
| 13 |
+
"std": [
|
| 14 |
+
2.0544580938943575
|
| 15 |
+
],
|
| 16 |
+
"count": [
|
| 17 |
+
7926
|
| 18 |
+
]
|
| 19 |
+
},
|
| 20 |
+
"timestamp": {
|
| 21 |
+
"min": [
|
| 22 |
+
0.0
|
| 23 |
+
],
|
| 24 |
+
"max": [
|
| 25 |
+
18.9
|
| 26 |
+
],
|
| 27 |
+
"mean": [
|
| 28 |
+
4.7648750946252845
|
| 29 |
+
],
|
| 30 |
+
"std": [
|
| 31 |
+
3.3602908404958978
|
| 32 |
+
],
|
| 33 |
+
"count": [
|
| 34 |
+
7926
|
| 35 |
+
]
|
| 36 |
+
},
|
| 37 |
+
"observation.images.main_camera": {
|
| 38 |
+
"min": [
|
| 39 |
+
[
|
| 40 |
+
[
|
| 41 |
+
0.0
|
| 42 |
+
]
|
| 43 |
+
],
|
| 44 |
+
[
|
| 45 |
+
[
|
| 46 |
+
0.0
|
| 47 |
+
]
|
| 48 |
+
],
|
| 49 |
+
[
|
| 50 |
+
[
|
| 51 |
+
0.0
|
| 52 |
+
]
|
| 53 |
+
]
|
| 54 |
+
],
|
| 55 |
+
"max": [
|
| 56 |
+
[
|
| 57 |
+
[
|
| 58 |
+
1.0
|
| 59 |
+
]
|
| 60 |
+
],
|
| 61 |
+
[
|
| 62 |
+
[
|
| 63 |
+
1.0
|
| 64 |
+
]
|
| 65 |
+
],
|
| 66 |
+
[
|
| 67 |
+
[
|
| 68 |
+
1.0
|
| 69 |
+
]
|
| 70 |
+
]
|
| 71 |
+
],
|
| 72 |
+
"mean": [
|
| 73 |
+
[
|
| 74 |
+
[
|
| 75 |
+
0.555528060079293
|
| 76 |
+
]
|
| 77 |
+
],
|
| 78 |
+
[
|
| 79 |
+
[
|
| 80 |
+
0.5001672667560041
|
| 81 |
+
]
|
| 82 |
+
],
|
| 83 |
+
[
|
| 84 |
+
[
|
| 85 |
+
0.4661934710659206
|
| 86 |
+
]
|
| 87 |
+
]
|
| 88 |
+
],
|
| 89 |
+
"std": [
|
| 90 |
+
[
|
| 91 |
+
[
|
| 92 |
+
0.27726402390679156
|
| 93 |
+
]
|
| 94 |
+
],
|
| 95 |
+
[
|
| 96 |
+
[
|
| 97 |
+
0.2692959828344183
|
| 98 |
+
]
|
| 99 |
+
],
|
| 100 |
+
[
|
| 101 |
+
[
|
| 102 |
+
0.29166607042108417
|
| 103 |
+
]
|
| 104 |
+
]
|
| 105 |
+
],
|
| 106 |
+
"count": [
|
| 107 |
+
7327
|
| 108 |
+
]
|
| 109 |
+
},
|
| 110 |
+
"frame_index": {
|
| 111 |
+
"min": [
|
| 112 |
+
0
|
| 113 |
+
],
|
| 114 |
+
"max": [
|
| 115 |
+
189
|
| 116 |
+
],
|
| 117 |
+
"mean": [
|
| 118 |
+
47.64875094625284
|
| 119 |
+
],
|
| 120 |
+
"std": [
|
| 121 |
+
33.60290840495898
|
| 122 |
+
],
|
| 123 |
+
"count": [
|
| 124 |
+
7926
|
| 125 |
+
]
|
| 126 |
+
},
|
| 127 |
+
"episode_index": {
|
| 128 |
+
"min": [
|
| 129 |
+
0
|
| 130 |
+
],
|
| 131 |
+
"max": [
|
| 132 |
+
93
|
| 133 |
+
],
|
| 134 |
+
"mean": [
|
| 135 |
+
45.76507696189755
|
| 136 |
+
],
|
| 137 |
+
"std": [
|
| 138 |
+
25.492871983974286
|
| 139 |
+
],
|
| 140 |
+
"count": [
|
| 141 |
+
7926
|
| 142 |
+
]
|
| 143 |
+
},
|
| 144 |
+
"action": {
|
| 145 |
+
"min": [
|
| 146 |
+
-0.009999999776482582,
|
| 147 |
+
-0.009999999776482582,
|
| 148 |
+
-0.009999999776482582,
|
| 149 |
+
-0.06283185631036758,
|
| 150 |
+
-0.06283185631036758,
|
| 151 |
+
-0.06283185631036758,
|
| 152 |
+
0.0
|
| 153 |
+
],
|
| 154 |
+
"max": [
|
| 155 |
+
0.009999999776482582,
|
| 156 |
+
0.009999999776482582,
|
| 157 |
+
0.009999999776482582,
|
| 158 |
+
0.06283185631036758,
|
| 159 |
+
0.06283185631036758,
|
| 160 |
+
0.06283185631036758,
|
| 161 |
+
1.0
|
| 162 |
+
],
|
| 163 |
+
"mean": [
|
| 164 |
+
-0.0007279836889540412,
|
| 165 |
+
-0.0006913952369976209,
|
| 166 |
+
-0.0014635374462693043,
|
| 167 |
+
0.0001506188750439639,
|
| 168 |
+
0.0005945482963482033,
|
| 169 |
+
0.00043600208816539665,
|
| 170 |
+
0.19782992727056226
|
| 171 |
+
],
|
| 172 |
+
"std": [
|
| 173 |
+
0.00456123754324318,
|
| 174 |
+
0.004601329721796265,
|
| 175 |
+
0.006874959713725188,
|
| 176 |
+
0.006349991607626829,
|
| 177 |
+
0.010037842266341796,
|
| 178 |
+
0.022678942702988626,
|
| 179 |
+
0.3983632021101758
|
| 180 |
+
],
|
| 181 |
+
"count": [
|
| 182 |
+
7926
|
| 183 |
+
],
|
| 184 |
+
"mask": [
|
| 185 |
+
true,
|
| 186 |
+
true,
|
| 187 |
+
true,
|
| 188 |
+
true,
|
| 189 |
+
true,
|
| 190 |
+
true,
|
| 191 |
+
false
|
| 192 |
+
]
|
| 193 |
+
},
|
| 194 |
+
"index": {
|
| 195 |
+
"min": [
|
| 196 |
+
0
|
| 197 |
+
],
|
| 198 |
+
"max": [
|
| 199 |
+
7925
|
| 200 |
+
],
|
| 201 |
+
"mean": [
|
| 202 |
+
3962.5
|
| 203 |
+
],
|
| 204 |
+
"std": [
|
| 205 |
+
2288.0390985878425
|
| 206 |
+
],
|
| 207 |
+
"count": [
|
| 208 |
+
7926
|
| 209 |
+
]
|
| 210 |
+
},
|
| 211 |
+
"observation.state": {
|
| 212 |
+
"min": [
|
| 213 |
+
-0.7127864360809326,
|
| 214 |
+
-0.26552391052246094,
|
| 215 |
+
-0.22021731734275818,
|
| 216 |
+
-3.141585111618042,
|
| 217 |
+
-1.213579535484314,
|
| 218 |
+
-3.129558563232422,
|
| 219 |
+
0.0
|
| 220 |
+
],
|
| 221 |
+
"max": [
|
| 222 |
+
-0.18012529611587524,
|
| 223 |
+
0.40225377678871155,
|
| 224 |
+
0.3791935443878174,
|
| 225 |
+
3.141571283340454,
|
| 226 |
+
0.7122786045074463,
|
| 227 |
+
3.139193534851074,
|
| 228 |
+
1.0
|
| 229 |
+
],
|
| 230 |
+
"mean": [
|
| 231 |
+
-0.4996030880192329,
|
| 232 |
+
0.009509113360888263,
|
| 233 |
+
-0.00720238650948428,
|
| 234 |
+
0.11769092076021163,
|
| 235 |
+
-0.13661677501693842,
|
| 236 |
+
-0.9702019779349543,
|
| 237 |
+
0.18622256048332025
|
| 238 |
+
],
|
| 239 |
+
"std": [
|
| 240 |
+
0.09823799387042138,
|
| 241 |
+
0.10558754729804731,
|
| 242 |
+
0.08036906002359806,
|
| 243 |
+
3.024165750594254,
|
| 244 |
+
0.30352495618631314,
|
| 245 |
+
1.3630765131948437,
|
| 246 |
+
0.3892861724640222
|
| 247 |
+
],
|
| 248 |
+
"count": [
|
| 249 |
+
7926
|
| 250 |
+
]
|
| 251 |
+
}
|
| 252 |
+
}
|
| 253 |
+
]
|
test-5-checkpoint-500/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:26bb84f4cc7a5913d60998c4f2fff465b32b9c8ba132266c44c82aeb662dc4ee
|
| 3 |
+
size 404630680
|
test-5-checkpoint-500/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e0c680800641a70932197b071dfe00cf9f8c08e376482ef7a405f208568ba197
|
| 3 |
+
size 29305786
|
test-5-checkpoint-500/rng_state_0.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2d8016e2b1257358774d31b654f3502f80624967b983d047290bb9bfc78adf1b
|
| 3 |
+
size 15984
|
test-5-checkpoint-500/rng_state_1.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7d0c9bbbda8d4bc60c8602cdbcbb21cf26f8c91c28833a25aa2d0a7d702cd0e5
|
| 3 |
+
size 15984
|
test-5-checkpoint-500/rng_state_2.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0cafb3d2f77805c07e34e3d61e3cc4e0b89b26a98adcf0752a6493b898f5146c
|
| 3 |
+
size 15984
|
test-5-checkpoint-500/rng_state_3.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8663a08d2bdeb011ef2a76ea2f9ee52973680d941df1bcfb7ac02ef30e9a1991
|
| 3 |
+
size 15984
|
test-5-checkpoint-500/rng_state_4.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e06346d051508c3a7f8ec11c561db450d60ff7dffcb567b59710ae4a01885353
|
| 3 |
+
size 15984
|
test-5-checkpoint-500/rng_state_5.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a7c34ce08a2cd7d7b60bc7403233e5b3760aedcc88615e706aab2e3a484c7308
|
| 3 |
+
size 15984
|
test-5-checkpoint-500/rng_state_6.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ead1ce276c3dbbcf038f17b11c67b983b931a59e449110d900bd7f9d05457d0f
|
| 3 |
+
size 15984
|
test-5-checkpoint-500/rng_state_7.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b043a4343e6d70b87a371a1bfd2ec6caeb955c206b6699367afbd373e03cd182
|
| 3 |
+
size 15984
|
test-5-checkpoint-500/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:722095417950d5a40eae3f57d8e53a42def012f8bda7ae218d2ae2f6092e373c
|
| 3 |
+
size 1064
|
test-5-checkpoint-500/trainer_state.json
ADDED
|
@@ -0,0 +1,108 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.4,
|
| 5 |
+
"eval_steps": 100,
|
| 6 |
+
"global_step": 500,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.08,
|
| 13 |
+
"grad_norm": 3.543903112411499,
|
| 14 |
+
"learning_rate": 5e-05,
|
| 15 |
+
"loss": 0.4585,
|
| 16 |
+
"step": 100
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.08,
|
| 20 |
+
"eval_loss": 0.3906002640724182,
|
| 21 |
+
"eval_runtime": 3.4549,
|
| 22 |
+
"eval_samples_per_second": 1095.823,
|
| 23 |
+
"eval_steps_per_second": 137.195,
|
| 24 |
+
"step": 100
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.16,
|
| 28 |
+
"grad_norm": 2.947075128555298,
|
| 29 |
+
"learning_rate": 5e-05,
|
| 30 |
+
"loss": 0.297,
|
| 31 |
+
"step": 200
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.16,
|
| 35 |
+
"eval_loss": 0.3242250680923462,
|
| 36 |
+
"eval_runtime": 2.9547,
|
| 37 |
+
"eval_samples_per_second": 1281.363,
|
| 38 |
+
"eval_steps_per_second": 160.424,
|
| 39 |
+
"step": 200
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.24,
|
| 43 |
+
"grad_norm": 3.019486665725708,
|
| 44 |
+
"learning_rate": 5e-05,
|
| 45 |
+
"loss": 0.261,
|
| 46 |
+
"step": 300
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 0.24,
|
| 50 |
+
"eval_loss": 0.3360763490200043,
|
| 51 |
+
"eval_runtime": 2.8774,
|
| 52 |
+
"eval_samples_per_second": 1315.769,
|
| 53 |
+
"eval_steps_per_second": 164.732,
|
| 54 |
+
"step": 300
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 0.32,
|
| 58 |
+
"grad_norm": 3.393603563308716,
|
| 59 |
+
"learning_rate": 5e-05,
|
| 60 |
+
"loss": 0.2395,
|
| 61 |
+
"step": 400
|
| 62 |
+
},
|
| 63 |
+
{
|
| 64 |
+
"epoch": 0.32,
|
| 65 |
+
"eval_loss": 0.3030490279197693,
|
| 66 |
+
"eval_runtime": 2.8913,
|
| 67 |
+
"eval_samples_per_second": 1309.46,
|
| 68 |
+
"eval_steps_per_second": 163.942,
|
| 69 |
+
"step": 400
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.4,
|
| 73 |
+
"grad_norm": 2.703859329223633,
|
| 74 |
+
"learning_rate": 5e-05,
|
| 75 |
+
"loss": 0.2263,
|
| 76 |
+
"step": 500
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 0.4,
|
| 80 |
+
"eval_loss": 0.2896880507469177,
|
| 81 |
+
"eval_runtime": 2.8735,
|
| 82 |
+
"eval_samples_per_second": 1317.54,
|
| 83 |
+
"eval_steps_per_second": 164.954,
|
| 84 |
+
"step": 500
|
| 85 |
+
}
|
| 86 |
+
],
|
| 87 |
+
"logging_steps": 100,
|
| 88 |
+
"max_steps": 300000,
|
| 89 |
+
"num_input_tokens_seen": 0,
|
| 90 |
+
"num_train_epochs": 240,
|
| 91 |
+
"save_steps": 500,
|
| 92 |
+
"stateful_callbacks": {
|
| 93 |
+
"TrainerControl": {
|
| 94 |
+
"args": {
|
| 95 |
+
"should_epoch_stop": false,
|
| 96 |
+
"should_evaluate": false,
|
| 97 |
+
"should_log": false,
|
| 98 |
+
"should_save": true,
|
| 99 |
+
"should_training_stop": false
|
| 100 |
+
},
|
| 101 |
+
"attributes": {}
|
| 102 |
+
}
|
| 103 |
+
},
|
| 104 |
+
"total_flos": 0.0,
|
| 105 |
+
"train_batch_size": 1,
|
| 106 |
+
"trial_name": null,
|
| 107 |
+
"trial_params": null
|
| 108 |
+
}
|
test-5-checkpoint-500/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5af203d1e14f5f535e1e4598a6c1bd50f6c174b130f9c5233ea55d2672400d54
|
| 3 |
+
size 5496
|