Upload folder using huggingface_hub
Browse files- checkpoint-10000/config.json +29 -0
- checkpoint-10000/generation_config.json +7 -0
- checkpoint-10000/model.safetensors +3 -0
- checkpoint-10000/optimizer.pt +3 -0
- checkpoint-10000/rng_state.pth +3 -0
- checkpoint-10000/scheduler.pt +3 -0
- checkpoint-10000/trainer_state.json +733 -0
- checkpoint-10000/training_args.bin +3 -0
- checkpoint-5000/config.json +29 -0
- checkpoint-5000/generation_config.json +7 -0
- checkpoint-5000/model.safetensors +3 -0
- checkpoint-5000/optimizer.pt +3 -0
- checkpoint-5000/rng_state.pth +3 -0
- checkpoint-5000/scheduler.pt +3 -0
- checkpoint-5000/trainer_state.json +383 -0
- checkpoint-5000/training_args.bin +3 -0
- config.json +29 -0
- generation_config.json +7 -0
- model.safetensors +3 -0
- special_tokens_map.json +30 -0
- tokenizer.json +0 -0
- tokenizer_config.json +44 -0
- training_args.bin +3 -0
checkpoint-10000/config.json
ADDED
|
@@ -0,0 +1,29 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"T5ForConditionalGeneration"
|
| 4 |
+
],
|
| 5 |
+
"classifier_dropout": 0.0,
|
| 6 |
+
"d_ff": 2048,
|
| 7 |
+
"d_kv": 64,
|
| 8 |
+
"d_model": 512,
|
| 9 |
+
"decoder_start_token_id": 0,
|
| 10 |
+
"dense_act_fn": "relu",
|
| 11 |
+
"dropout_rate": 0.1,
|
| 12 |
+
"eos_token_id": 1,
|
| 13 |
+
"feed_forward_proj": "relu",
|
| 14 |
+
"initializer_factor": 1.0,
|
| 15 |
+
"is_encoder_decoder": true,
|
| 16 |
+
"is_gated_act": false,
|
| 17 |
+
"layer_norm_epsilon": 1e-06,
|
| 18 |
+
"model_type": "t5",
|
| 19 |
+
"num_decoder_layers": 6,
|
| 20 |
+
"num_heads": 8,
|
| 21 |
+
"num_layers": 6,
|
| 22 |
+
"pad_token_id": 0,
|
| 23 |
+
"relative_attention_max_distance": 128,
|
| 24 |
+
"relative_attention_num_buckets": 32,
|
| 25 |
+
"torch_dtype": "float32",
|
| 26 |
+
"transformers_version": "4.49.0",
|
| 27 |
+
"use_cache": true,
|
| 28 |
+
"vocab_size": 50000
|
| 29 |
+
}
|
checkpoint-10000/generation_config.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"decoder_start_token_id": 0,
|
| 4 |
+
"eos_token_id": 1,
|
| 5 |
+
"pad_token_id": 0,
|
| 6 |
+
"transformers_version": "4.49.0"
|
| 7 |
+
}
|
checkpoint-10000/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5b89923a516924762c2de248dacf4862b0e13476a487390a50a81cb29fc1bfe2
|
| 3 |
+
size 278643752
|
checkpoint-10000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:adb9858cb4b21765f0075a1542d9d3b1aac139fcff68be457f47f1ad7e1a1578
|
| 3 |
+
size 557367226
|
checkpoint-10000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f5a18679862da768a40e8967bb0dec5995b07cd29a8a6dc7221003e0843572bb
|
| 3 |
+
size 14244
|
checkpoint-10000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:23bacfc9e0420908692bc36f206e544fd539ad9548d49f18ad4437c98828bc0f
|
| 3 |
+
size 1064
|
checkpoint-10000/trainer_state.json
ADDED
|
@@ -0,0 +1,733 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 10000,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.01,
|
| 13 |
+
"grad_norm": 0.6850011944770813,
|
| 14 |
+
"learning_rate": 4.9500000000000004e-05,
|
| 15 |
+
"loss": 2.7276,
|
| 16 |
+
"step": 100
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.02,
|
| 20 |
+
"grad_norm": 1.3513778448104858,
|
| 21 |
+
"learning_rate": 4.9e-05,
|
| 22 |
+
"loss": 2.2015,
|
| 23 |
+
"step": 200
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.03,
|
| 27 |
+
"grad_norm": 1.4596431255340576,
|
| 28 |
+
"learning_rate": 4.85e-05,
|
| 29 |
+
"loss": 2.1439,
|
| 30 |
+
"step": 300
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.04,
|
| 34 |
+
"grad_norm": 1.4662152528762817,
|
| 35 |
+
"learning_rate": 4.8e-05,
|
| 36 |
+
"loss": 2.0443,
|
| 37 |
+
"step": 400
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.05,
|
| 41 |
+
"grad_norm": 1.5642160177230835,
|
| 42 |
+
"learning_rate": 4.75e-05,
|
| 43 |
+
"loss": 2.041,
|
| 44 |
+
"step": 500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.06,
|
| 48 |
+
"grad_norm": 1.5689924955368042,
|
| 49 |
+
"learning_rate": 4.7e-05,
|
| 50 |
+
"loss": 1.8782,
|
| 51 |
+
"step": 600
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.07,
|
| 55 |
+
"grad_norm": 1.754616141319275,
|
| 56 |
+
"learning_rate": 4.6500000000000005e-05,
|
| 57 |
+
"loss": 1.8376,
|
| 58 |
+
"step": 700
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.08,
|
| 62 |
+
"grad_norm": 1.4754304885864258,
|
| 63 |
+
"learning_rate": 4.600000000000001e-05,
|
| 64 |
+
"loss": 1.9273,
|
| 65 |
+
"step": 800
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.09,
|
| 69 |
+
"grad_norm": 1.6210129261016846,
|
| 70 |
+
"learning_rate": 4.55e-05,
|
| 71 |
+
"loss": 1.8568,
|
| 72 |
+
"step": 900
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.1,
|
| 76 |
+
"grad_norm": 1.5406711101531982,
|
| 77 |
+
"learning_rate": 4.5e-05,
|
| 78 |
+
"loss": 1.7329,
|
| 79 |
+
"step": 1000
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.11,
|
| 83 |
+
"grad_norm": 1.5473151206970215,
|
| 84 |
+
"learning_rate": 4.4500000000000004e-05,
|
| 85 |
+
"loss": 1.6887,
|
| 86 |
+
"step": 1100
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.12,
|
| 90 |
+
"grad_norm": 1.7042499780654907,
|
| 91 |
+
"learning_rate": 4.4000000000000006e-05,
|
| 92 |
+
"loss": 1.6909,
|
| 93 |
+
"step": 1200
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.13,
|
| 97 |
+
"grad_norm": 1.6643552780151367,
|
| 98 |
+
"learning_rate": 4.35e-05,
|
| 99 |
+
"loss": 1.7344,
|
| 100 |
+
"step": 1300
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.14,
|
| 104 |
+
"grad_norm": 1.515936255455017,
|
| 105 |
+
"learning_rate": 4.3e-05,
|
| 106 |
+
"loss": 1.6402,
|
| 107 |
+
"step": 1400
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.15,
|
| 111 |
+
"grad_norm": 1.430012583732605,
|
| 112 |
+
"learning_rate": 4.25e-05,
|
| 113 |
+
"loss": 1.595,
|
| 114 |
+
"step": 1500
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.16,
|
| 118 |
+
"grad_norm": 1.9956995248794556,
|
| 119 |
+
"learning_rate": 4.2e-05,
|
| 120 |
+
"loss": 1.5611,
|
| 121 |
+
"step": 1600
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.17,
|
| 125 |
+
"grad_norm": 1.8397706747055054,
|
| 126 |
+
"learning_rate": 4.15e-05,
|
| 127 |
+
"loss": 1.6238,
|
| 128 |
+
"step": 1700
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.18,
|
| 132 |
+
"grad_norm": 1.8368654251098633,
|
| 133 |
+
"learning_rate": 4.1e-05,
|
| 134 |
+
"loss": 1.5441,
|
| 135 |
+
"step": 1800
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.19,
|
| 139 |
+
"grad_norm": 1.6586986780166626,
|
| 140 |
+
"learning_rate": 4.05e-05,
|
| 141 |
+
"loss": 1.5188,
|
| 142 |
+
"step": 1900
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.2,
|
| 146 |
+
"grad_norm": 1.7311187982559204,
|
| 147 |
+
"learning_rate": 4e-05,
|
| 148 |
+
"loss": 1.516,
|
| 149 |
+
"step": 2000
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.21,
|
| 153 |
+
"grad_norm": 1.6303811073303223,
|
| 154 |
+
"learning_rate": 3.9500000000000005e-05,
|
| 155 |
+
"loss": 1.4869,
|
| 156 |
+
"step": 2100
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.22,
|
| 160 |
+
"grad_norm": 1.519205927848816,
|
| 161 |
+
"learning_rate": 3.9000000000000006e-05,
|
| 162 |
+
"loss": 1.4699,
|
| 163 |
+
"step": 2200
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.23,
|
| 167 |
+
"grad_norm": 1.656671404838562,
|
| 168 |
+
"learning_rate": 3.85e-05,
|
| 169 |
+
"loss": 1.4298,
|
| 170 |
+
"step": 2300
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.24,
|
| 174 |
+
"grad_norm": 1.5942782163619995,
|
| 175 |
+
"learning_rate": 3.8e-05,
|
| 176 |
+
"loss": 1.4659,
|
| 177 |
+
"step": 2400
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.25,
|
| 181 |
+
"grad_norm": 1.5101336240768433,
|
| 182 |
+
"learning_rate": 3.7500000000000003e-05,
|
| 183 |
+
"loss": 1.5001,
|
| 184 |
+
"step": 2500
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.26,
|
| 188 |
+
"grad_norm": 1.61963951587677,
|
| 189 |
+
"learning_rate": 3.7e-05,
|
| 190 |
+
"loss": 1.4175,
|
| 191 |
+
"step": 2600
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.27,
|
| 195 |
+
"grad_norm": 1.7214901447296143,
|
| 196 |
+
"learning_rate": 3.65e-05,
|
| 197 |
+
"loss": 1.402,
|
| 198 |
+
"step": 2700
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.28,
|
| 202 |
+
"grad_norm": 1.4414921998977661,
|
| 203 |
+
"learning_rate": 3.6e-05,
|
| 204 |
+
"loss": 1.4171,
|
| 205 |
+
"step": 2800
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.29,
|
| 209 |
+
"grad_norm": 1.7572053670883179,
|
| 210 |
+
"learning_rate": 3.55e-05,
|
| 211 |
+
"loss": 1.3965,
|
| 212 |
+
"step": 2900
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.3,
|
| 216 |
+
"grad_norm": 1.837003469467163,
|
| 217 |
+
"learning_rate": 3.5e-05,
|
| 218 |
+
"loss": 1.3753,
|
| 219 |
+
"step": 3000
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.31,
|
| 223 |
+
"grad_norm": 1.868739128112793,
|
| 224 |
+
"learning_rate": 3.45e-05,
|
| 225 |
+
"loss": 1.3463,
|
| 226 |
+
"step": 3100
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.32,
|
| 230 |
+
"grad_norm": 2.0180938243865967,
|
| 231 |
+
"learning_rate": 3.4000000000000007e-05,
|
| 232 |
+
"loss": 1.3115,
|
| 233 |
+
"step": 3200
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.33,
|
| 237 |
+
"grad_norm": 1.5846806764602661,
|
| 238 |
+
"learning_rate": 3.35e-05,
|
| 239 |
+
"loss": 1.3008,
|
| 240 |
+
"step": 3300
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.34,
|
| 244 |
+
"grad_norm": 1.7848007678985596,
|
| 245 |
+
"learning_rate": 3.3e-05,
|
| 246 |
+
"loss": 1.3129,
|
| 247 |
+
"step": 3400
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.35,
|
| 251 |
+
"grad_norm": 1.7048262357711792,
|
| 252 |
+
"learning_rate": 3.2500000000000004e-05,
|
| 253 |
+
"loss": 1.3294,
|
| 254 |
+
"step": 3500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.36,
|
| 258 |
+
"grad_norm": 1.9426958560943604,
|
| 259 |
+
"learning_rate": 3.2000000000000005e-05,
|
| 260 |
+
"loss": 1.2896,
|
| 261 |
+
"step": 3600
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.37,
|
| 265 |
+
"grad_norm": 2.048593521118164,
|
| 266 |
+
"learning_rate": 3.15e-05,
|
| 267 |
+
"loss": 1.2755,
|
| 268 |
+
"step": 3700
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.38,
|
| 272 |
+
"grad_norm": 1.6056958436965942,
|
| 273 |
+
"learning_rate": 3.1e-05,
|
| 274 |
+
"loss": 1.2465,
|
| 275 |
+
"step": 3800
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.39,
|
| 279 |
+
"grad_norm": 1.5590567588806152,
|
| 280 |
+
"learning_rate": 3.05e-05,
|
| 281 |
+
"loss": 1.2514,
|
| 282 |
+
"step": 3900
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.4,
|
| 286 |
+
"grad_norm": 1.8030258417129517,
|
| 287 |
+
"learning_rate": 3e-05,
|
| 288 |
+
"loss": 1.2493,
|
| 289 |
+
"step": 4000
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.41,
|
| 293 |
+
"grad_norm": 1.7520445585250854,
|
| 294 |
+
"learning_rate": 2.95e-05,
|
| 295 |
+
"loss": 1.2235,
|
| 296 |
+
"step": 4100
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.42,
|
| 300 |
+
"grad_norm": 1.6477694511413574,
|
| 301 |
+
"learning_rate": 2.9e-05,
|
| 302 |
+
"loss": 1.2036,
|
| 303 |
+
"step": 4200
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.43,
|
| 307 |
+
"grad_norm": 2.046046257019043,
|
| 308 |
+
"learning_rate": 2.8499999999999998e-05,
|
| 309 |
+
"loss": 1.2175,
|
| 310 |
+
"step": 4300
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.44,
|
| 314 |
+
"grad_norm": 1.769077181816101,
|
| 315 |
+
"learning_rate": 2.8000000000000003e-05,
|
| 316 |
+
"loss": 1.1884,
|
| 317 |
+
"step": 4400
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.45,
|
| 321 |
+
"grad_norm": 2.0415198802948,
|
| 322 |
+
"learning_rate": 2.7500000000000004e-05,
|
| 323 |
+
"loss": 1.2058,
|
| 324 |
+
"step": 4500
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.46,
|
| 328 |
+
"grad_norm": 1.9062659740447998,
|
| 329 |
+
"learning_rate": 2.7000000000000002e-05,
|
| 330 |
+
"loss": 1.1946,
|
| 331 |
+
"step": 4600
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.47,
|
| 335 |
+
"grad_norm": 2.0025410652160645,
|
| 336 |
+
"learning_rate": 2.6500000000000004e-05,
|
| 337 |
+
"loss": 1.1952,
|
| 338 |
+
"step": 4700
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.48,
|
| 342 |
+
"grad_norm": 1.717053771018982,
|
| 343 |
+
"learning_rate": 2.6000000000000002e-05,
|
| 344 |
+
"loss": 1.156,
|
| 345 |
+
"step": 4800
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.49,
|
| 349 |
+
"grad_norm": 1.6754528284072876,
|
| 350 |
+
"learning_rate": 2.5500000000000003e-05,
|
| 351 |
+
"loss": 1.1314,
|
| 352 |
+
"step": 4900
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.5,
|
| 356 |
+
"grad_norm": 1.3899461030960083,
|
| 357 |
+
"learning_rate": 2.5e-05,
|
| 358 |
+
"loss": 1.1576,
|
| 359 |
+
"step": 5000
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.51,
|
| 363 |
+
"grad_norm": 1.6504240036010742,
|
| 364 |
+
"learning_rate": 2.45e-05,
|
| 365 |
+
"loss": 1.1773,
|
| 366 |
+
"step": 5100
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.52,
|
| 370 |
+
"grad_norm": 1.95234215259552,
|
| 371 |
+
"learning_rate": 2.4e-05,
|
| 372 |
+
"loss": 1.1109,
|
| 373 |
+
"step": 5200
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.53,
|
| 377 |
+
"grad_norm": 1.9345238208770752,
|
| 378 |
+
"learning_rate": 2.35e-05,
|
| 379 |
+
"loss": 1.1648,
|
| 380 |
+
"step": 5300
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.54,
|
| 384 |
+
"grad_norm": 1.8829147815704346,
|
| 385 |
+
"learning_rate": 2.3000000000000003e-05,
|
| 386 |
+
"loss": 1.1723,
|
| 387 |
+
"step": 5400
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.55,
|
| 391 |
+
"grad_norm": 1.8588093519210815,
|
| 392 |
+
"learning_rate": 2.25e-05,
|
| 393 |
+
"loss": 1.1159,
|
| 394 |
+
"step": 5500
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.56,
|
| 398 |
+
"grad_norm": 1.7027171850204468,
|
| 399 |
+
"learning_rate": 2.2000000000000003e-05,
|
| 400 |
+
"loss": 1.0798,
|
| 401 |
+
"step": 5600
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.57,
|
| 405 |
+
"grad_norm": 1.7587237358093262,
|
| 406 |
+
"learning_rate": 2.15e-05,
|
| 407 |
+
"loss": 1.1179,
|
| 408 |
+
"step": 5700
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.58,
|
| 412 |
+
"grad_norm": 2.0505311489105225,
|
| 413 |
+
"learning_rate": 2.1e-05,
|
| 414 |
+
"loss": 1.1142,
|
| 415 |
+
"step": 5800
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.59,
|
| 419 |
+
"grad_norm": 2.03979229927063,
|
| 420 |
+
"learning_rate": 2.05e-05,
|
| 421 |
+
"loss": 1.1376,
|
| 422 |
+
"step": 5900
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.6,
|
| 426 |
+
"grad_norm": 1.6683259010314941,
|
| 427 |
+
"learning_rate": 2e-05,
|
| 428 |
+
"loss": 1.1018,
|
| 429 |
+
"step": 6000
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.61,
|
| 433 |
+
"grad_norm": 2.0077784061431885,
|
| 434 |
+
"learning_rate": 1.9500000000000003e-05,
|
| 435 |
+
"loss": 1.1112,
|
| 436 |
+
"step": 6100
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.62,
|
| 440 |
+
"grad_norm": 2.2150514125823975,
|
| 441 |
+
"learning_rate": 1.9e-05,
|
| 442 |
+
"loss": 1.0573,
|
| 443 |
+
"step": 6200
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.63,
|
| 447 |
+
"grad_norm": 1.886559247970581,
|
| 448 |
+
"learning_rate": 1.85e-05,
|
| 449 |
+
"loss": 1.0841,
|
| 450 |
+
"step": 6300
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.64,
|
| 454 |
+
"grad_norm": 2.1064834594726562,
|
| 455 |
+
"learning_rate": 1.8e-05,
|
| 456 |
+
"loss": 1.0877,
|
| 457 |
+
"step": 6400
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.65,
|
| 461 |
+
"grad_norm": 1.7919803857803345,
|
| 462 |
+
"learning_rate": 1.75e-05,
|
| 463 |
+
"loss": 1.1124,
|
| 464 |
+
"step": 6500
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.66,
|
| 468 |
+
"grad_norm": 1.6519834995269775,
|
| 469 |
+
"learning_rate": 1.7000000000000003e-05,
|
| 470 |
+
"loss": 1.0847,
|
| 471 |
+
"step": 6600
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.67,
|
| 475 |
+
"grad_norm": 2.4856765270233154,
|
| 476 |
+
"learning_rate": 1.65e-05,
|
| 477 |
+
"loss": 1.0531,
|
| 478 |
+
"step": 6700
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.68,
|
| 482 |
+
"grad_norm": 2.2001657485961914,
|
| 483 |
+
"learning_rate": 1.6000000000000003e-05,
|
| 484 |
+
"loss": 1.0623,
|
| 485 |
+
"step": 6800
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.69,
|
| 489 |
+
"grad_norm": 1.8861461877822876,
|
| 490 |
+
"learning_rate": 1.55e-05,
|
| 491 |
+
"loss": 1.066,
|
| 492 |
+
"step": 6900
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.7,
|
| 496 |
+
"grad_norm": 2.00785756111145,
|
| 497 |
+
"learning_rate": 1.5e-05,
|
| 498 |
+
"loss": 1.0618,
|
| 499 |
+
"step": 7000
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.71,
|
| 503 |
+
"grad_norm": 2.0481674671173096,
|
| 504 |
+
"learning_rate": 1.45e-05,
|
| 505 |
+
"loss": 1.0417,
|
| 506 |
+
"step": 7100
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.72,
|
| 510 |
+
"grad_norm": 1.885738492012024,
|
| 511 |
+
"learning_rate": 1.4000000000000001e-05,
|
| 512 |
+
"loss": 1.0537,
|
| 513 |
+
"step": 7200
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.73,
|
| 517 |
+
"grad_norm": 2.0609591007232666,
|
| 518 |
+
"learning_rate": 1.3500000000000001e-05,
|
| 519 |
+
"loss": 1.0298,
|
| 520 |
+
"step": 7300
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.74,
|
| 524 |
+
"grad_norm": 1.767872929573059,
|
| 525 |
+
"learning_rate": 1.3000000000000001e-05,
|
| 526 |
+
"loss": 1.0427,
|
| 527 |
+
"step": 7400
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.75,
|
| 531 |
+
"grad_norm": 1.8356056213378906,
|
| 532 |
+
"learning_rate": 1.25e-05,
|
| 533 |
+
"loss": 1.0745,
|
| 534 |
+
"step": 7500
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.76,
|
| 538 |
+
"grad_norm": 1.8087313175201416,
|
| 539 |
+
"learning_rate": 1.2e-05,
|
| 540 |
+
"loss": 1.0031,
|
| 541 |
+
"step": 7600
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.77,
|
| 545 |
+
"grad_norm": 2.0896551609039307,
|
| 546 |
+
"learning_rate": 1.1500000000000002e-05,
|
| 547 |
+
"loss": 1.04,
|
| 548 |
+
"step": 7700
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.78,
|
| 552 |
+
"grad_norm": 2.0215108394622803,
|
| 553 |
+
"learning_rate": 1.1000000000000001e-05,
|
| 554 |
+
"loss": 1.0274,
|
| 555 |
+
"step": 7800
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.79,
|
| 559 |
+
"grad_norm": 1.8265337944030762,
|
| 560 |
+
"learning_rate": 1.05e-05,
|
| 561 |
+
"loss": 1.0075,
|
| 562 |
+
"step": 7900
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.8,
|
| 566 |
+
"grad_norm": 1.8919756412506104,
|
| 567 |
+
"learning_rate": 1e-05,
|
| 568 |
+
"loss": 1.0197,
|
| 569 |
+
"step": 8000
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.81,
|
| 573 |
+
"grad_norm": 1.8041898012161255,
|
| 574 |
+
"learning_rate": 9.5e-06,
|
| 575 |
+
"loss": 1.0187,
|
| 576 |
+
"step": 8100
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.82,
|
| 580 |
+
"grad_norm": 2.347409963607788,
|
| 581 |
+
"learning_rate": 9e-06,
|
| 582 |
+
"loss": 1.0439,
|
| 583 |
+
"step": 8200
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.83,
|
| 587 |
+
"grad_norm": 2.4638195037841797,
|
| 588 |
+
"learning_rate": 8.500000000000002e-06,
|
| 589 |
+
"loss": 1.0234,
|
| 590 |
+
"step": 8300
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.84,
|
| 594 |
+
"grad_norm": 2.1736338138580322,
|
| 595 |
+
"learning_rate": 8.000000000000001e-06,
|
| 596 |
+
"loss": 1.0265,
|
| 597 |
+
"step": 8400
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.85,
|
| 601 |
+
"grad_norm": 2.583214282989502,
|
| 602 |
+
"learning_rate": 7.5e-06,
|
| 603 |
+
"loss": 1.0163,
|
| 604 |
+
"step": 8500
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.86,
|
| 608 |
+
"grad_norm": 2.069798707962036,
|
| 609 |
+
"learning_rate": 7.000000000000001e-06,
|
| 610 |
+
"loss": 1.0214,
|
| 611 |
+
"step": 8600
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.87,
|
| 615 |
+
"grad_norm": 2.100463628768921,
|
| 616 |
+
"learning_rate": 6.5000000000000004e-06,
|
| 617 |
+
"loss": 1.0216,
|
| 618 |
+
"step": 8700
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.88,
|
| 622 |
+
"grad_norm": 1.9746285676956177,
|
| 623 |
+
"learning_rate": 6e-06,
|
| 624 |
+
"loss": 0.9824,
|
| 625 |
+
"step": 8800
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.89,
|
| 629 |
+
"grad_norm": 2.1781320571899414,
|
| 630 |
+
"learning_rate": 5.500000000000001e-06,
|
| 631 |
+
"loss": 1.022,
|
| 632 |
+
"step": 8900
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.9,
|
| 636 |
+
"grad_norm": 1.9690595865249634,
|
| 637 |
+
"learning_rate": 5e-06,
|
| 638 |
+
"loss": 0.979,
|
| 639 |
+
"step": 9000
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.91,
|
| 643 |
+
"grad_norm": 2.073143243789673,
|
| 644 |
+
"learning_rate": 4.5e-06,
|
| 645 |
+
"loss": 1.002,
|
| 646 |
+
"step": 9100
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.92,
|
| 650 |
+
"grad_norm": 2.178834915161133,
|
| 651 |
+
"learning_rate": 4.000000000000001e-06,
|
| 652 |
+
"loss": 1.0046,
|
| 653 |
+
"step": 9200
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.93,
|
| 657 |
+
"grad_norm": 1.87750244140625,
|
| 658 |
+
"learning_rate": 3.5000000000000004e-06,
|
| 659 |
+
"loss": 1.0171,
|
| 660 |
+
"step": 9300
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.94,
|
| 664 |
+
"grad_norm": 1.8957669734954834,
|
| 665 |
+
"learning_rate": 3e-06,
|
| 666 |
+
"loss": 1.004,
|
| 667 |
+
"step": 9400
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.95,
|
| 671 |
+
"grad_norm": 2.0873405933380127,
|
| 672 |
+
"learning_rate": 2.5e-06,
|
| 673 |
+
"loss": 1.0296,
|
| 674 |
+
"step": 9500
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.96,
|
| 678 |
+
"grad_norm": 1.8929296731948853,
|
| 679 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 680 |
+
"loss": 0.981,
|
| 681 |
+
"step": 9600
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.97,
|
| 685 |
+
"grad_norm": 2.0239217281341553,
|
| 686 |
+
"learning_rate": 1.5e-06,
|
| 687 |
+
"loss": 0.9793,
|
| 688 |
+
"step": 9700
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.98,
|
| 692 |
+
"grad_norm": 1.8227028846740723,
|
| 693 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 694 |
+
"loss": 0.9962,
|
| 695 |
+
"step": 9800
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.99,
|
| 699 |
+
"grad_norm": 1.9512608051300049,
|
| 700 |
+
"learning_rate": 5.000000000000001e-07,
|
| 701 |
+
"loss": 0.982,
|
| 702 |
+
"step": 9900
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 1.0,
|
| 706 |
+
"grad_norm": 2.1475226879119873,
|
| 707 |
+
"learning_rate": 0.0,
|
| 708 |
+
"loss": 1.0236,
|
| 709 |
+
"step": 10000
|
| 710 |
+
}
|
| 711 |
+
],
|
| 712 |
+
"logging_steps": 100,
|
| 713 |
+
"max_steps": 10000,
|
| 714 |
+
"num_input_tokens_seen": 0,
|
| 715 |
+
"num_train_epochs": 9223372036854775807,
|
| 716 |
+
"save_steps": 5000,
|
| 717 |
+
"stateful_callbacks": {
|
| 718 |
+
"TrainerControl": {
|
| 719 |
+
"args": {
|
| 720 |
+
"should_epoch_stop": false,
|
| 721 |
+
"should_evaluate": false,
|
| 722 |
+
"should_log": false,
|
| 723 |
+
"should_save": true,
|
| 724 |
+
"should_training_stop": true
|
| 725 |
+
},
|
| 726 |
+
"attributes": {}
|
| 727 |
+
}
|
| 728 |
+
},
|
| 729 |
+
"total_flos": 5413672058880000.0,
|
| 730 |
+
"train_batch_size": 32,
|
| 731 |
+
"trial_name": null,
|
| 732 |
+
"trial_params": null
|
| 733 |
+
}
|
checkpoint-10000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f4db73719714726f2f557eac6d22a8fa80f1a235536b2aea4ec20c0b19b14dbe
|
| 3 |
+
size 5432
|
checkpoint-5000/config.json
ADDED
|
@@ -0,0 +1,29 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"T5ForConditionalGeneration"
|
| 4 |
+
],
|
| 5 |
+
"classifier_dropout": 0.0,
|
| 6 |
+
"d_ff": 2048,
|
| 7 |
+
"d_kv": 64,
|
| 8 |
+
"d_model": 512,
|
| 9 |
+
"decoder_start_token_id": 0,
|
| 10 |
+
"dense_act_fn": "relu",
|
| 11 |
+
"dropout_rate": 0.1,
|
| 12 |
+
"eos_token_id": 1,
|
| 13 |
+
"feed_forward_proj": "relu",
|
| 14 |
+
"initializer_factor": 1.0,
|
| 15 |
+
"is_encoder_decoder": true,
|
| 16 |
+
"is_gated_act": false,
|
| 17 |
+
"layer_norm_epsilon": 1e-06,
|
| 18 |
+
"model_type": "t5",
|
| 19 |
+
"num_decoder_layers": 6,
|
| 20 |
+
"num_heads": 8,
|
| 21 |
+
"num_layers": 6,
|
| 22 |
+
"pad_token_id": 0,
|
| 23 |
+
"relative_attention_max_distance": 128,
|
| 24 |
+
"relative_attention_num_buckets": 32,
|
| 25 |
+
"torch_dtype": "float32",
|
| 26 |
+
"transformers_version": "4.49.0",
|
| 27 |
+
"use_cache": true,
|
| 28 |
+
"vocab_size": 50000
|
| 29 |
+
}
|
checkpoint-5000/generation_config.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"decoder_start_token_id": 0,
|
| 4 |
+
"eos_token_id": 1,
|
| 5 |
+
"pad_token_id": 0,
|
| 6 |
+
"transformers_version": "4.49.0"
|
| 7 |
+
}
|
checkpoint-5000/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4eeb3aac5a4db5e4cdd17be1c73f12e6c2df4585422b306817d7e4a5117cc1ed
|
| 3 |
+
size 278643752
|
checkpoint-5000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:113f49eabe05f7b39e29b887c1fb8fb80ccac228936098155e55cb5d58517cd8
|
| 3 |
+
size 557367226
|
checkpoint-5000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d8d179c379a56f256ef3d5cc887682b1d2236d40085a156b1779080652f583fc
|
| 3 |
+
size 14244
|
checkpoint-5000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a78c8c8a4be6cb8ff4b8d2d9805ee7c29168ba8bd60ae33f3adc872613cd729f
|
| 3 |
+
size 1064
|
checkpoint-5000/trainer_state.json
ADDED
|
@@ -0,0 +1,383 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.5,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 5000,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.01,
|
| 13 |
+
"grad_norm": 0.6850011944770813,
|
| 14 |
+
"learning_rate": 4.9500000000000004e-05,
|
| 15 |
+
"loss": 2.7276,
|
| 16 |
+
"step": 100
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.02,
|
| 20 |
+
"grad_norm": 1.3513778448104858,
|
| 21 |
+
"learning_rate": 4.9e-05,
|
| 22 |
+
"loss": 2.2015,
|
| 23 |
+
"step": 200
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.03,
|
| 27 |
+
"grad_norm": 1.4596431255340576,
|
| 28 |
+
"learning_rate": 4.85e-05,
|
| 29 |
+
"loss": 2.1439,
|
| 30 |
+
"step": 300
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.04,
|
| 34 |
+
"grad_norm": 1.4662152528762817,
|
| 35 |
+
"learning_rate": 4.8e-05,
|
| 36 |
+
"loss": 2.0443,
|
| 37 |
+
"step": 400
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.05,
|
| 41 |
+
"grad_norm": 1.5642160177230835,
|
| 42 |
+
"learning_rate": 4.75e-05,
|
| 43 |
+
"loss": 2.041,
|
| 44 |
+
"step": 500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.06,
|
| 48 |
+
"grad_norm": 1.5689924955368042,
|
| 49 |
+
"learning_rate": 4.7e-05,
|
| 50 |
+
"loss": 1.8782,
|
| 51 |
+
"step": 600
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.07,
|
| 55 |
+
"grad_norm": 1.754616141319275,
|
| 56 |
+
"learning_rate": 4.6500000000000005e-05,
|
| 57 |
+
"loss": 1.8376,
|
| 58 |
+
"step": 700
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.08,
|
| 62 |
+
"grad_norm": 1.4754304885864258,
|
| 63 |
+
"learning_rate": 4.600000000000001e-05,
|
| 64 |
+
"loss": 1.9273,
|
| 65 |
+
"step": 800
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.09,
|
| 69 |
+
"grad_norm": 1.6210129261016846,
|
| 70 |
+
"learning_rate": 4.55e-05,
|
| 71 |
+
"loss": 1.8568,
|
| 72 |
+
"step": 900
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.1,
|
| 76 |
+
"grad_norm": 1.5406711101531982,
|
| 77 |
+
"learning_rate": 4.5e-05,
|
| 78 |
+
"loss": 1.7329,
|
| 79 |
+
"step": 1000
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.11,
|
| 83 |
+
"grad_norm": 1.5473151206970215,
|
| 84 |
+
"learning_rate": 4.4500000000000004e-05,
|
| 85 |
+
"loss": 1.6887,
|
| 86 |
+
"step": 1100
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.12,
|
| 90 |
+
"grad_norm": 1.7042499780654907,
|
| 91 |
+
"learning_rate": 4.4000000000000006e-05,
|
| 92 |
+
"loss": 1.6909,
|
| 93 |
+
"step": 1200
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.13,
|
| 97 |
+
"grad_norm": 1.6643552780151367,
|
| 98 |
+
"learning_rate": 4.35e-05,
|
| 99 |
+
"loss": 1.7344,
|
| 100 |
+
"step": 1300
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.14,
|
| 104 |
+
"grad_norm": 1.515936255455017,
|
| 105 |
+
"learning_rate": 4.3e-05,
|
| 106 |
+
"loss": 1.6402,
|
| 107 |
+
"step": 1400
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.15,
|
| 111 |
+
"grad_norm": 1.430012583732605,
|
| 112 |
+
"learning_rate": 4.25e-05,
|
| 113 |
+
"loss": 1.595,
|
| 114 |
+
"step": 1500
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.16,
|
| 118 |
+
"grad_norm": 1.9956995248794556,
|
| 119 |
+
"learning_rate": 4.2e-05,
|
| 120 |
+
"loss": 1.5611,
|
| 121 |
+
"step": 1600
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.17,
|
| 125 |
+
"grad_norm": 1.8397706747055054,
|
| 126 |
+
"learning_rate": 4.15e-05,
|
| 127 |
+
"loss": 1.6238,
|
| 128 |
+
"step": 1700
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.18,
|
| 132 |
+
"grad_norm": 1.8368654251098633,
|
| 133 |
+
"learning_rate": 4.1e-05,
|
| 134 |
+
"loss": 1.5441,
|
| 135 |
+
"step": 1800
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.19,
|
| 139 |
+
"grad_norm": 1.6586986780166626,
|
| 140 |
+
"learning_rate": 4.05e-05,
|
| 141 |
+
"loss": 1.5188,
|
| 142 |
+
"step": 1900
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.2,
|
| 146 |
+
"grad_norm": 1.7311187982559204,
|
| 147 |
+
"learning_rate": 4e-05,
|
| 148 |
+
"loss": 1.516,
|
| 149 |
+
"step": 2000
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.21,
|
| 153 |
+
"grad_norm": 1.6303811073303223,
|
| 154 |
+
"learning_rate": 3.9500000000000005e-05,
|
| 155 |
+
"loss": 1.4869,
|
| 156 |
+
"step": 2100
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.22,
|
| 160 |
+
"grad_norm": 1.519205927848816,
|
| 161 |
+
"learning_rate": 3.9000000000000006e-05,
|
| 162 |
+
"loss": 1.4699,
|
| 163 |
+
"step": 2200
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.23,
|
| 167 |
+
"grad_norm": 1.656671404838562,
|
| 168 |
+
"learning_rate": 3.85e-05,
|
| 169 |
+
"loss": 1.4298,
|
| 170 |
+
"step": 2300
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.24,
|
| 174 |
+
"grad_norm": 1.5942782163619995,
|
| 175 |
+
"learning_rate": 3.8e-05,
|
| 176 |
+
"loss": 1.4659,
|
| 177 |
+
"step": 2400
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.25,
|
| 181 |
+
"grad_norm": 1.5101336240768433,
|
| 182 |
+
"learning_rate": 3.7500000000000003e-05,
|
| 183 |
+
"loss": 1.5001,
|
| 184 |
+
"step": 2500
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.26,
|
| 188 |
+
"grad_norm": 1.61963951587677,
|
| 189 |
+
"learning_rate": 3.7e-05,
|
| 190 |
+
"loss": 1.4175,
|
| 191 |
+
"step": 2600
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.27,
|
| 195 |
+
"grad_norm": 1.7214901447296143,
|
| 196 |
+
"learning_rate": 3.65e-05,
|
| 197 |
+
"loss": 1.402,
|
| 198 |
+
"step": 2700
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.28,
|
| 202 |
+
"grad_norm": 1.4414921998977661,
|
| 203 |
+
"learning_rate": 3.6e-05,
|
| 204 |
+
"loss": 1.4171,
|
| 205 |
+
"step": 2800
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.29,
|
| 209 |
+
"grad_norm": 1.7572053670883179,
|
| 210 |
+
"learning_rate": 3.55e-05,
|
| 211 |
+
"loss": 1.3965,
|
| 212 |
+
"step": 2900
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.3,
|
| 216 |
+
"grad_norm": 1.837003469467163,
|
| 217 |
+
"learning_rate": 3.5e-05,
|
| 218 |
+
"loss": 1.3753,
|
| 219 |
+
"step": 3000
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.31,
|
| 223 |
+
"grad_norm": 1.868739128112793,
|
| 224 |
+
"learning_rate": 3.45e-05,
|
| 225 |
+
"loss": 1.3463,
|
| 226 |
+
"step": 3100
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.32,
|
| 230 |
+
"grad_norm": 2.0180938243865967,
|
| 231 |
+
"learning_rate": 3.4000000000000007e-05,
|
| 232 |
+
"loss": 1.3115,
|
| 233 |
+
"step": 3200
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.33,
|
| 237 |
+
"grad_norm": 1.5846806764602661,
|
| 238 |
+
"learning_rate": 3.35e-05,
|
| 239 |
+
"loss": 1.3008,
|
| 240 |
+
"step": 3300
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.34,
|
| 244 |
+
"grad_norm": 1.7848007678985596,
|
| 245 |
+
"learning_rate": 3.3e-05,
|
| 246 |
+
"loss": 1.3129,
|
| 247 |
+
"step": 3400
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.35,
|
| 251 |
+
"grad_norm": 1.7048262357711792,
|
| 252 |
+
"learning_rate": 3.2500000000000004e-05,
|
| 253 |
+
"loss": 1.3294,
|
| 254 |
+
"step": 3500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.36,
|
| 258 |
+
"grad_norm": 1.9426958560943604,
|
| 259 |
+
"learning_rate": 3.2000000000000005e-05,
|
| 260 |
+
"loss": 1.2896,
|
| 261 |
+
"step": 3600
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.37,
|
| 265 |
+
"grad_norm": 2.048593521118164,
|
| 266 |
+
"learning_rate": 3.15e-05,
|
| 267 |
+
"loss": 1.2755,
|
| 268 |
+
"step": 3700
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.38,
|
| 272 |
+
"grad_norm": 1.6056958436965942,
|
| 273 |
+
"learning_rate": 3.1e-05,
|
| 274 |
+
"loss": 1.2465,
|
| 275 |
+
"step": 3800
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.39,
|
| 279 |
+
"grad_norm": 1.5590567588806152,
|
| 280 |
+
"learning_rate": 3.05e-05,
|
| 281 |
+
"loss": 1.2514,
|
| 282 |
+
"step": 3900
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.4,
|
| 286 |
+
"grad_norm": 1.8030258417129517,
|
| 287 |
+
"learning_rate": 3e-05,
|
| 288 |
+
"loss": 1.2493,
|
| 289 |
+
"step": 4000
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.41,
|
| 293 |
+
"grad_norm": 1.7520445585250854,
|
| 294 |
+
"learning_rate": 2.95e-05,
|
| 295 |
+
"loss": 1.2235,
|
| 296 |
+
"step": 4100
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.42,
|
| 300 |
+
"grad_norm": 1.6477694511413574,
|
| 301 |
+
"learning_rate": 2.9e-05,
|
| 302 |
+
"loss": 1.2036,
|
| 303 |
+
"step": 4200
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.43,
|
| 307 |
+
"grad_norm": 2.046046257019043,
|
| 308 |
+
"learning_rate": 2.8499999999999998e-05,
|
| 309 |
+
"loss": 1.2175,
|
| 310 |
+
"step": 4300
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.44,
|
| 314 |
+
"grad_norm": 1.769077181816101,
|
| 315 |
+
"learning_rate": 2.8000000000000003e-05,
|
| 316 |
+
"loss": 1.1884,
|
| 317 |
+
"step": 4400
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.45,
|
| 321 |
+
"grad_norm": 2.0415198802948,
|
| 322 |
+
"learning_rate": 2.7500000000000004e-05,
|
| 323 |
+
"loss": 1.2058,
|
| 324 |
+
"step": 4500
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.46,
|
| 328 |
+
"grad_norm": 1.9062659740447998,
|
| 329 |
+
"learning_rate": 2.7000000000000002e-05,
|
| 330 |
+
"loss": 1.1946,
|
| 331 |
+
"step": 4600
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.47,
|
| 335 |
+
"grad_norm": 2.0025410652160645,
|
| 336 |
+
"learning_rate": 2.6500000000000004e-05,
|
| 337 |
+
"loss": 1.1952,
|
| 338 |
+
"step": 4700
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.48,
|
| 342 |
+
"grad_norm": 1.717053771018982,
|
| 343 |
+
"learning_rate": 2.6000000000000002e-05,
|
| 344 |
+
"loss": 1.156,
|
| 345 |
+
"step": 4800
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.49,
|
| 349 |
+
"grad_norm": 1.6754528284072876,
|
| 350 |
+
"learning_rate": 2.5500000000000003e-05,
|
| 351 |
+
"loss": 1.1314,
|
| 352 |
+
"step": 4900
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.5,
|
| 356 |
+
"grad_norm": 1.3899461030960083,
|
| 357 |
+
"learning_rate": 2.5e-05,
|
| 358 |
+
"loss": 1.1576,
|
| 359 |
+
"step": 5000
|
| 360 |
+
}
|
| 361 |
+
],
|
| 362 |
+
"logging_steps": 100,
|
| 363 |
+
"max_steps": 10000,
|
| 364 |
+
"num_input_tokens_seen": 0,
|
| 365 |
+
"num_train_epochs": 9223372036854775807,
|
| 366 |
+
"save_steps": 5000,
|
| 367 |
+
"stateful_callbacks": {
|
| 368 |
+
"TrainerControl": {
|
| 369 |
+
"args": {
|
| 370 |
+
"should_epoch_stop": false,
|
| 371 |
+
"should_evaluate": false,
|
| 372 |
+
"should_log": false,
|
| 373 |
+
"should_save": true,
|
| 374 |
+
"should_training_stop": false
|
| 375 |
+
},
|
| 376 |
+
"attributes": {}
|
| 377 |
+
}
|
| 378 |
+
},
|
| 379 |
+
"total_flos": 2706836029440000.0,
|
| 380 |
+
"train_batch_size": 32,
|
| 381 |
+
"trial_name": null,
|
| 382 |
+
"trial_params": null
|
| 383 |
+
}
|
checkpoint-5000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f4db73719714726f2f557eac6d22a8fa80f1a235536b2aea4ec20c0b19b14dbe
|
| 3 |
+
size 5432
|
config.json
ADDED
|
@@ -0,0 +1,29 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"T5ForConditionalGeneration"
|
| 4 |
+
],
|
| 5 |
+
"classifier_dropout": 0.0,
|
| 6 |
+
"d_ff": 2048,
|
| 7 |
+
"d_kv": 64,
|
| 8 |
+
"d_model": 512,
|
| 9 |
+
"decoder_start_token_id": 0,
|
| 10 |
+
"dense_act_fn": "relu",
|
| 11 |
+
"dropout_rate": 0.1,
|
| 12 |
+
"eos_token_id": 1,
|
| 13 |
+
"feed_forward_proj": "relu",
|
| 14 |
+
"initializer_factor": 1.0,
|
| 15 |
+
"is_encoder_decoder": true,
|
| 16 |
+
"is_gated_act": false,
|
| 17 |
+
"layer_norm_epsilon": 1e-06,
|
| 18 |
+
"model_type": "t5",
|
| 19 |
+
"num_decoder_layers": 6,
|
| 20 |
+
"num_heads": 8,
|
| 21 |
+
"num_layers": 6,
|
| 22 |
+
"pad_token_id": 0,
|
| 23 |
+
"relative_attention_max_distance": 128,
|
| 24 |
+
"relative_attention_num_buckets": 32,
|
| 25 |
+
"torch_dtype": "float32",
|
| 26 |
+
"transformers_version": "4.49.0",
|
| 27 |
+
"use_cache": true,
|
| 28 |
+
"vocab_size": 50000
|
| 29 |
+
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"decoder_start_token_id": 0,
|
| 4 |
+
"eos_token_id": 1,
|
| 5 |
+
"pad_token_id": 0,
|
| 6 |
+
"transformers_version": "4.49.0"
|
| 7 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5b89923a516924762c2de248dacf4862b0e13476a487390a50a81cb29fc1bfe2
|
| 3 |
+
size 278643752
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"eos_token": {
|
| 10 |
+
"content": "</s>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"pad_token": {
|
| 17 |
+
"content": "<pad>",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": false,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
},
|
| 23 |
+
"unk_token": {
|
| 24 |
+
"content": "<unk>",
|
| 25 |
+
"lstrip": false,
|
| 26 |
+
"normalized": false,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
}
|
| 30 |
+
}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,44 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "<pad>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<unk>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "<s>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"3": {
|
| 28 |
+
"content": "</s>",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
}
|
| 35 |
+
},
|
| 36 |
+
"bos_token": "<s>",
|
| 37 |
+
"clean_up_tokenization_spaces": false,
|
| 38 |
+
"eos_token": "</s>",
|
| 39 |
+
"extra_special_tokens": {},
|
| 40 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 41 |
+
"pad_token": "<pad>",
|
| 42 |
+
"tokenizer_class": "PreTrainedTokenizer",
|
| 43 |
+
"unk_token": "<unk>"
|
| 44 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f4db73719714726f2f557eac6d22a8fa80f1a235536b2aea4ec20c0b19b14dbe
|
| 3 |
+
size 5432
|