initial commit, for 12-layer gpt2-like transformer (checkpoint 27500)
Browse files- config.json +30 -0
- optimizer.pt +3 -0
- pytorch_model.bin +3 -0
- rng_state.pth +3 -0
- scaler.pt +3 -0
- scheduler.pt +3 -0
- trainer_state.json +731 -0
- training_args.bin +3 -0
config.json
ADDED
|
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"activation_function": "gelu_new",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"GPT2LMHeadModel"
|
| 5 |
+
],
|
| 6 |
+
"attn_pdrop": 0.1,
|
| 7 |
+
"bos_token_id": 0,
|
| 8 |
+
"embd_pdrop": 0.1,
|
| 9 |
+
"eos_token_id": 0,
|
| 10 |
+
"gradient_checkpointing": false,
|
| 11 |
+
"initializer_range": 0.02,
|
| 12 |
+
"layer_norm_epsilon": 1e-05,
|
| 13 |
+
"model_type": "gpt2",
|
| 14 |
+
"n_ctx": 1024,
|
| 15 |
+
"n_embd": 768,
|
| 16 |
+
"n_head": 12,
|
| 17 |
+
"n_inner": null,
|
| 18 |
+
"n_layer": 12,
|
| 19 |
+
"n_positions": 1024,
|
| 20 |
+
"resid_pdrop": 0.1,
|
| 21 |
+
"scale_attn_weights": true,
|
| 22 |
+
"summary_activation": null,
|
| 23 |
+
"summary_first_dropout": 0.1,
|
| 24 |
+
"summary_proj_to_labels": true,
|
| 25 |
+
"summary_type": "cls_index",
|
| 26 |
+
"summary_use_proj": true,
|
| 27 |
+
"transformers_version": "4.6.1",
|
| 28 |
+
"use_cache": true,
|
| 29 |
+
"vocab_size": 28439
|
| 30 |
+
}
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:62c1070d05b3d1b91fc0bc1b97da9e54e059a45f7209d886647fdc9bf5e06e44
|
| 3 |
+
size 861554225
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5790dfe2444d9065db58e118e3005e2073a006a6fa412d7a29e41e57859969eb
|
| 3 |
+
size 443378921
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:943be73215e3a61b82175bf421f3436f19bdb2b1f531d7c1ce24aa2b9f455822
|
| 3 |
+
size 14503
|
scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8ea0883070082a76534d38eab2641c971d0073a8b8f47b3f84bfd52eb4b7cf00
|
| 3 |
+
size 559
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f756cb62bd490582930035cc4e9b15c2f2efe7a2e83d3c252a602413a8a2e9bc
|
| 3 |
+
size 623
|
trainer_state.json
ADDED
|
@@ -0,0 +1,731 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 3.7010114192962646,
|
| 3 |
+
"best_model_checkpoint": "/scratch/ka2773/project/lm-mem/checkpoints/gpt2_full_12-768-1024_a_01/checkpoint-27500",
|
| 4 |
+
"epoch": 2.9165340969349876,
|
| 5 |
+
"global_step": 27500,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.05,
|
| 12 |
+
"learning_rate": 6e-05,
|
| 13 |
+
"loss": 6.6816,
|
| 14 |
+
"step": 500
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.05,
|
| 18 |
+
"eval_loss": 6.0154523849487305,
|
| 19 |
+
"eval_runtime": 9.6363,
|
| 20 |
+
"eval_samples_per_second": 24.491,
|
| 21 |
+
"step": 500
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.11,
|
| 25 |
+
"learning_rate": 6e-05,
|
| 26 |
+
"loss": 6.014,
|
| 27 |
+
"step": 1000
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 0.11,
|
| 31 |
+
"eval_loss": 5.75693941116333,
|
| 32 |
+
"eval_runtime": 9.6373,
|
| 33 |
+
"eval_samples_per_second": 24.488,
|
| 34 |
+
"step": 1000
|
| 35 |
+
},
|
| 36 |
+
{
|
| 37 |
+
"epoch": 0.16,
|
| 38 |
+
"learning_rate": 6e-05,
|
| 39 |
+
"loss": 5.8347,
|
| 40 |
+
"step": 1500
|
| 41 |
+
},
|
| 42 |
+
{
|
| 43 |
+
"epoch": 0.16,
|
| 44 |
+
"eval_loss": 5.60043478012085,
|
| 45 |
+
"eval_runtime": 9.6586,
|
| 46 |
+
"eval_samples_per_second": 24.434,
|
| 47 |
+
"step": 1500
|
| 48 |
+
},
|
| 49 |
+
{
|
| 50 |
+
"epoch": 0.21,
|
| 51 |
+
"learning_rate": 6e-05,
|
| 52 |
+
"loss": 5.6946,
|
| 53 |
+
"step": 2000
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 0.21,
|
| 57 |
+
"eval_loss": 5.451482772827148,
|
| 58 |
+
"eval_runtime": 9.6385,
|
| 59 |
+
"eval_samples_per_second": 24.485,
|
| 60 |
+
"step": 2000
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 0.27,
|
| 64 |
+
"learning_rate": 6e-05,
|
| 65 |
+
"loss": 5.5628,
|
| 66 |
+
"step": 2500
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.27,
|
| 70 |
+
"eval_loss": 5.31380558013916,
|
| 71 |
+
"eval_runtime": 9.6551,
|
| 72 |
+
"eval_samples_per_second": 24.443,
|
| 73 |
+
"step": 2500
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.32,
|
| 77 |
+
"learning_rate": 6e-05,
|
| 78 |
+
"loss": 5.4328,
|
| 79 |
+
"step": 3000
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.32,
|
| 83 |
+
"eval_loss": 5.174571990966797,
|
| 84 |
+
"eval_runtime": 9.6558,
|
| 85 |
+
"eval_samples_per_second": 24.441,
|
| 86 |
+
"step": 3000
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.37,
|
| 90 |
+
"learning_rate": 6e-05,
|
| 91 |
+
"loss": 5.3018,
|
| 92 |
+
"step": 3500
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.37,
|
| 96 |
+
"eval_loss": 5.04353666305542,
|
| 97 |
+
"eval_runtime": 9.6566,
|
| 98 |
+
"eval_samples_per_second": 24.439,
|
| 99 |
+
"step": 3500
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.42,
|
| 103 |
+
"learning_rate": 6e-05,
|
| 104 |
+
"loss": 5.1768,
|
| 105 |
+
"step": 4000
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.42,
|
| 109 |
+
"eval_loss": 4.923619270324707,
|
| 110 |
+
"eval_runtime": 9.6443,
|
| 111 |
+
"eval_samples_per_second": 24.47,
|
| 112 |
+
"step": 4000
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 0.48,
|
| 116 |
+
"learning_rate": 6e-05,
|
| 117 |
+
"loss": 5.0737,
|
| 118 |
+
"step": 4500
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"epoch": 0.48,
|
| 122 |
+
"eval_loss": 4.811069011688232,
|
| 123 |
+
"eval_runtime": 9.6397,
|
| 124 |
+
"eval_samples_per_second": 24.482,
|
| 125 |
+
"step": 4500
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"epoch": 0.53,
|
| 129 |
+
"learning_rate": 6e-05,
|
| 130 |
+
"loss": 4.9689,
|
| 131 |
+
"step": 5000
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 0.53,
|
| 135 |
+
"eval_loss": 4.7268242835998535,
|
| 136 |
+
"eval_runtime": 9.6562,
|
| 137 |
+
"eval_samples_per_second": 24.44,
|
| 138 |
+
"step": 5000
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 0.58,
|
| 142 |
+
"learning_rate": 6e-05,
|
| 143 |
+
"loss": 4.8992,
|
| 144 |
+
"step": 5500
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 0.58,
|
| 148 |
+
"eval_loss": 4.646885871887207,
|
| 149 |
+
"eval_runtime": 9.6343,
|
| 150 |
+
"eval_samples_per_second": 24.496,
|
| 151 |
+
"step": 5500
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 0.64,
|
| 155 |
+
"learning_rate": 6e-05,
|
| 156 |
+
"loss": 4.8284,
|
| 157 |
+
"step": 6000
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.64,
|
| 161 |
+
"eval_loss": 4.5805535316467285,
|
| 162 |
+
"eval_runtime": 9.6391,
|
| 163 |
+
"eval_samples_per_second": 24.484,
|
| 164 |
+
"step": 6000
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.69,
|
| 168 |
+
"learning_rate": 6e-05,
|
| 169 |
+
"loss": 4.77,
|
| 170 |
+
"step": 6500
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.69,
|
| 174 |
+
"eval_loss": 4.520021915435791,
|
| 175 |
+
"eval_runtime": 9.6435,
|
| 176 |
+
"eval_samples_per_second": 24.472,
|
| 177 |
+
"step": 6500
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.74,
|
| 181 |
+
"learning_rate": 6e-05,
|
| 182 |
+
"loss": 4.7052,
|
| 183 |
+
"step": 7000
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 0.74,
|
| 187 |
+
"eval_loss": 4.462948322296143,
|
| 188 |
+
"eval_runtime": 9.6368,
|
| 189 |
+
"eval_samples_per_second": 24.489,
|
| 190 |
+
"step": 7000
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 0.8,
|
| 194 |
+
"learning_rate": 6e-05,
|
| 195 |
+
"loss": 4.6502,
|
| 196 |
+
"step": 7500
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 0.8,
|
| 200 |
+
"eval_loss": 4.413123607635498,
|
| 201 |
+
"eval_runtime": 9.6344,
|
| 202 |
+
"eval_samples_per_second": 24.496,
|
| 203 |
+
"step": 7500
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 0.85,
|
| 207 |
+
"learning_rate": 6e-05,
|
| 208 |
+
"loss": 4.6089,
|
| 209 |
+
"step": 8000
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 0.85,
|
| 213 |
+
"eval_loss": 4.364911079406738,
|
| 214 |
+
"eval_runtime": 9.6314,
|
| 215 |
+
"eval_samples_per_second": 24.503,
|
| 216 |
+
"step": 8000
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 0.9,
|
| 220 |
+
"learning_rate": 6e-05,
|
| 221 |
+
"loss": 4.5489,
|
| 222 |
+
"step": 8500
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 0.9,
|
| 226 |
+
"eval_loss": 4.311524391174316,
|
| 227 |
+
"eval_runtime": 9.6277,
|
| 228 |
+
"eval_samples_per_second": 24.513,
|
| 229 |
+
"step": 8500
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"epoch": 0.95,
|
| 233 |
+
"learning_rate": 6e-05,
|
| 234 |
+
"loss": 4.5042,
|
| 235 |
+
"step": 9000
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 0.95,
|
| 239 |
+
"eval_loss": 4.259948253631592,
|
| 240 |
+
"eval_runtime": 9.6594,
|
| 241 |
+
"eval_samples_per_second": 24.432,
|
| 242 |
+
"step": 9000
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 1.01,
|
| 246 |
+
"learning_rate": 6e-05,
|
| 247 |
+
"loss": 4.454,
|
| 248 |
+
"step": 9500
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 1.01,
|
| 252 |
+
"eval_loss": 4.214750289916992,
|
| 253 |
+
"eval_runtime": 9.5772,
|
| 254 |
+
"eval_samples_per_second": 24.642,
|
| 255 |
+
"step": 9500
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 1.06,
|
| 259 |
+
"learning_rate": 6e-05,
|
| 260 |
+
"loss": 4.3863,
|
| 261 |
+
"step": 10000
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 1.06,
|
| 265 |
+
"eval_loss": 4.168990612030029,
|
| 266 |
+
"eval_runtime": 9.5953,
|
| 267 |
+
"eval_samples_per_second": 24.595,
|
| 268 |
+
"step": 10000
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 1.11,
|
| 272 |
+
"learning_rate": 6e-05,
|
| 273 |
+
"loss": 4.3451,
|
| 274 |
+
"step": 10500
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 1.11,
|
| 278 |
+
"eval_loss": 4.131111145019531,
|
| 279 |
+
"eval_runtime": 9.5871,
|
| 280 |
+
"eval_samples_per_second": 24.616,
|
| 281 |
+
"step": 10500
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 1.17,
|
| 285 |
+
"learning_rate": 6e-05,
|
| 286 |
+
"loss": 4.3209,
|
| 287 |
+
"step": 11000
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 1.17,
|
| 291 |
+
"eval_loss": 4.101424217224121,
|
| 292 |
+
"eval_runtime": 9.5799,
|
| 293 |
+
"eval_samples_per_second": 24.635,
|
| 294 |
+
"step": 11000
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 1.22,
|
| 298 |
+
"learning_rate": 6e-05,
|
| 299 |
+
"loss": 4.2959,
|
| 300 |
+
"step": 11500
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 1.22,
|
| 304 |
+
"eval_loss": 4.075887203216553,
|
| 305 |
+
"eval_runtime": 9.5788,
|
| 306 |
+
"eval_samples_per_second": 24.638,
|
| 307 |
+
"step": 11500
|
| 308 |
+
},
|
| 309 |
+
{
|
| 310 |
+
"epoch": 1.27,
|
| 311 |
+
"learning_rate": 6e-05,
|
| 312 |
+
"loss": 4.259,
|
| 313 |
+
"step": 12000
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 1.27,
|
| 317 |
+
"eval_loss": 4.046448230743408,
|
| 318 |
+
"eval_runtime": 9.5828,
|
| 319 |
+
"eval_samples_per_second": 24.628,
|
| 320 |
+
"step": 12000
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 1.33,
|
| 324 |
+
"learning_rate": 6e-05,
|
| 325 |
+
"loss": 4.2412,
|
| 326 |
+
"step": 12500
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 1.33,
|
| 330 |
+
"eval_loss": 4.025355339050293,
|
| 331 |
+
"eval_runtime": 9.5605,
|
| 332 |
+
"eval_samples_per_second": 24.685,
|
| 333 |
+
"step": 12500
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 1.38,
|
| 337 |
+
"learning_rate": 6e-05,
|
| 338 |
+
"loss": 4.2107,
|
| 339 |
+
"step": 13000
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 1.38,
|
| 343 |
+
"eval_loss": 4.008790493011475,
|
| 344 |
+
"eval_runtime": 9.5742,
|
| 345 |
+
"eval_samples_per_second": 24.65,
|
| 346 |
+
"step": 13000
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 1.43,
|
| 350 |
+
"learning_rate": 6e-05,
|
| 351 |
+
"loss": 4.1955,
|
| 352 |
+
"step": 13500
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 1.43,
|
| 356 |
+
"eval_loss": 3.980191707611084,
|
| 357 |
+
"eval_runtime": 9.5885,
|
| 358 |
+
"eval_samples_per_second": 24.613,
|
| 359 |
+
"step": 13500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.48,
|
| 363 |
+
"learning_rate": 6e-05,
|
| 364 |
+
"loss": 4.1816,
|
| 365 |
+
"step": 14000
|
| 366 |
+
},
|
| 367 |
+
{
|
| 368 |
+
"epoch": 1.48,
|
| 369 |
+
"eval_loss": 3.955965995788574,
|
| 370 |
+
"eval_runtime": 9.5843,
|
| 371 |
+
"eval_samples_per_second": 24.624,
|
| 372 |
+
"step": 14000
|
| 373 |
+
},
|
| 374 |
+
{
|
| 375 |
+
"epoch": 1.54,
|
| 376 |
+
"learning_rate": 6e-05,
|
| 377 |
+
"loss": 4.1469,
|
| 378 |
+
"step": 14500
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 1.54,
|
| 382 |
+
"eval_loss": 3.9338672161102295,
|
| 383 |
+
"eval_runtime": 9.5686,
|
| 384 |
+
"eval_samples_per_second": 24.664,
|
| 385 |
+
"step": 14500
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 1.59,
|
| 389 |
+
"learning_rate": 6e-05,
|
| 390 |
+
"loss": 4.1401,
|
| 391 |
+
"step": 15000
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 1.59,
|
| 395 |
+
"eval_loss": 3.91983962059021,
|
| 396 |
+
"eval_runtime": 9.5696,
|
| 397 |
+
"eval_samples_per_second": 24.661,
|
| 398 |
+
"step": 15000
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 1.64,
|
| 402 |
+
"learning_rate": 6e-05,
|
| 403 |
+
"loss": 4.1208,
|
| 404 |
+
"step": 15500
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 1.64,
|
| 408 |
+
"eval_loss": 3.906564712524414,
|
| 409 |
+
"eval_runtime": 9.5628,
|
| 410 |
+
"eval_samples_per_second": 24.679,
|
| 411 |
+
"step": 15500
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 1.7,
|
| 415 |
+
"learning_rate": 6e-05,
|
| 416 |
+
"loss": 4.1191,
|
| 417 |
+
"step": 16000
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 1.7,
|
| 421 |
+
"eval_loss": 3.89436674118042,
|
| 422 |
+
"eval_runtime": 9.5623,
|
| 423 |
+
"eval_samples_per_second": 24.68,
|
| 424 |
+
"step": 16000
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 1.75,
|
| 428 |
+
"learning_rate": 6e-05,
|
| 429 |
+
"loss": 4.0999,
|
| 430 |
+
"step": 16500
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 1.75,
|
| 434 |
+
"eval_loss": 3.884828805923462,
|
| 435 |
+
"eval_runtime": 9.5239,
|
| 436 |
+
"eval_samples_per_second": 24.78,
|
| 437 |
+
"step": 16500
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 1.8,
|
| 441 |
+
"learning_rate": 6e-05,
|
| 442 |
+
"loss": 4.09,
|
| 443 |
+
"step": 17000
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 1.8,
|
| 447 |
+
"eval_loss": 3.877413034439087,
|
| 448 |
+
"eval_runtime": 9.5641,
|
| 449 |
+
"eval_samples_per_second": 24.675,
|
| 450 |
+
"step": 17000
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 1.86,
|
| 454 |
+
"learning_rate": 6e-05,
|
| 455 |
+
"loss": 4.084,
|
| 456 |
+
"step": 17500
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 1.86,
|
| 460 |
+
"eval_loss": 3.8598787784576416,
|
| 461 |
+
"eval_runtime": 9.5727,
|
| 462 |
+
"eval_samples_per_second": 24.653,
|
| 463 |
+
"step": 17500
|
| 464 |
+
},
|
| 465 |
+
{
|
| 466 |
+
"epoch": 1.91,
|
| 467 |
+
"learning_rate": 6e-05,
|
| 468 |
+
"loss": 4.0619,
|
| 469 |
+
"step": 18000
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 1.91,
|
| 473 |
+
"eval_loss": 3.843931198120117,
|
| 474 |
+
"eval_runtime": 9.5583,
|
| 475 |
+
"eval_samples_per_second": 24.691,
|
| 476 |
+
"step": 18000
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 1.96,
|
| 480 |
+
"learning_rate": 6e-05,
|
| 481 |
+
"loss": 4.049,
|
| 482 |
+
"step": 18500
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 1.96,
|
| 486 |
+
"eval_loss": 3.8279693126678467,
|
| 487 |
+
"eval_runtime": 9.5719,
|
| 488 |
+
"eval_samples_per_second": 24.656,
|
| 489 |
+
"step": 18500
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 2.02,
|
| 493 |
+
"learning_rate": 6e-05,
|
| 494 |
+
"loss": 4.0287,
|
| 495 |
+
"step": 19000
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 2.02,
|
| 499 |
+
"eval_loss": 3.817969799041748,
|
| 500 |
+
"eval_runtime": 9.5289,
|
| 501 |
+
"eval_samples_per_second": 24.767,
|
| 502 |
+
"step": 19000
|
| 503 |
+
},
|
| 504 |
+
{
|
| 505 |
+
"epoch": 2.07,
|
| 506 |
+
"learning_rate": 6e-05,
|
| 507 |
+
"loss": 4.0008,
|
| 508 |
+
"step": 19500
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 2.07,
|
| 512 |
+
"eval_loss": 3.8063206672668457,
|
| 513 |
+
"eval_runtime": 9.5741,
|
| 514 |
+
"eval_samples_per_second": 24.65,
|
| 515 |
+
"step": 19500
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 2.12,
|
| 519 |
+
"learning_rate": 6e-05,
|
| 520 |
+
"loss": 3.9993,
|
| 521 |
+
"step": 20000
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 2.12,
|
| 525 |
+
"eval_loss": 3.80794358253479,
|
| 526 |
+
"eval_runtime": 9.5702,
|
| 527 |
+
"eval_samples_per_second": 24.66,
|
| 528 |
+
"step": 20000
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 2.17,
|
| 532 |
+
"learning_rate": 6e-05,
|
| 533 |
+
"loss": 3.9836,
|
| 534 |
+
"step": 20500
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 2.17,
|
| 538 |
+
"eval_loss": 3.7982654571533203,
|
| 539 |
+
"eval_runtime": 9.5716,
|
| 540 |
+
"eval_samples_per_second": 24.656,
|
| 541 |
+
"step": 20500
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 2.23,
|
| 545 |
+
"learning_rate": 6e-05,
|
| 546 |
+
"loss": 3.9813,
|
| 547 |
+
"step": 21000
|
| 548 |
+
},
|
| 549 |
+
{
|
| 550 |
+
"epoch": 2.23,
|
| 551 |
+
"eval_loss": 3.7907371520996094,
|
| 552 |
+
"eval_runtime": 9.5685,
|
| 553 |
+
"eval_samples_per_second": 24.664,
|
| 554 |
+
"step": 21000
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 2.28,
|
| 558 |
+
"learning_rate": 6e-05,
|
| 559 |
+
"loss": 3.9785,
|
| 560 |
+
"step": 21500
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 2.28,
|
| 564 |
+
"eval_loss": 3.7781858444213867,
|
| 565 |
+
"eval_runtime": 9.5717,
|
| 566 |
+
"eval_samples_per_second": 24.656,
|
| 567 |
+
"step": 21500
|
| 568 |
+
},
|
| 569 |
+
{
|
| 570 |
+
"epoch": 2.33,
|
| 571 |
+
"learning_rate": 6e-05,
|
| 572 |
+
"loss": 3.9704,
|
| 573 |
+
"step": 22000
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 2.33,
|
| 577 |
+
"eval_loss": 3.772528648376465,
|
| 578 |
+
"eval_runtime": 9.5802,
|
| 579 |
+
"eval_samples_per_second": 24.634,
|
| 580 |
+
"step": 22000
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"epoch": 2.39,
|
| 584 |
+
"learning_rate": 6e-05,
|
| 585 |
+
"loss": 3.9741,
|
| 586 |
+
"step": 22500
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 2.39,
|
| 590 |
+
"eval_loss": 3.778170108795166,
|
| 591 |
+
"eval_runtime": 9.5981,
|
| 592 |
+
"eval_samples_per_second": 24.588,
|
| 593 |
+
"step": 22500
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 2.44,
|
| 597 |
+
"learning_rate": 6e-05,
|
| 598 |
+
"loss": 3.9587,
|
| 599 |
+
"step": 23000
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 2.44,
|
| 603 |
+
"eval_loss": 3.7567873001098633,
|
| 604 |
+
"eval_runtime": 9.5943,
|
| 605 |
+
"eval_samples_per_second": 24.598,
|
| 606 |
+
"step": 23000
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 2.49,
|
| 610 |
+
"learning_rate": 6e-05,
|
| 611 |
+
"loss": 3.9505,
|
| 612 |
+
"step": 23500
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 2.49,
|
| 616 |
+
"eval_loss": 3.7593204975128174,
|
| 617 |
+
"eval_runtime": 9.589,
|
| 618 |
+
"eval_samples_per_second": 24.611,
|
| 619 |
+
"step": 23500
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 2.55,
|
| 623 |
+
"learning_rate": 6e-05,
|
| 624 |
+
"loss": 3.9358,
|
| 625 |
+
"step": 24000
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 2.55,
|
| 629 |
+
"eval_loss": 3.750127077102661,
|
| 630 |
+
"eval_runtime": 9.5871,
|
| 631 |
+
"eval_samples_per_second": 24.616,
|
| 632 |
+
"step": 24000
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 2.6,
|
| 636 |
+
"learning_rate": 6e-05,
|
| 637 |
+
"loss": 3.9371,
|
| 638 |
+
"step": 24500
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 2.6,
|
| 642 |
+
"eval_loss": 3.7427163124084473,
|
| 643 |
+
"eval_runtime": 9.5925,
|
| 644 |
+
"eval_samples_per_second": 24.602,
|
| 645 |
+
"step": 24500
|
| 646 |
+
},
|
| 647 |
+
{
|
| 648 |
+
"epoch": 2.65,
|
| 649 |
+
"learning_rate": 6e-05,
|
| 650 |
+
"loss": 3.9379,
|
| 651 |
+
"step": 25000
|
| 652 |
+
},
|
| 653 |
+
{
|
| 654 |
+
"epoch": 2.65,
|
| 655 |
+
"eval_loss": 3.7363016605377197,
|
| 656 |
+
"eval_runtime": 9.5892,
|
| 657 |
+
"eval_samples_per_second": 24.611,
|
| 658 |
+
"step": 25000
|
| 659 |
+
},
|
| 660 |
+
{
|
| 661 |
+
"epoch": 2.7,
|
| 662 |
+
"learning_rate": 6e-05,
|
| 663 |
+
"loss": 3.9352,
|
| 664 |
+
"step": 25500
|
| 665 |
+
},
|
| 666 |
+
{
|
| 667 |
+
"epoch": 2.7,
|
| 668 |
+
"eval_loss": 3.7250401973724365,
|
| 669 |
+
"eval_runtime": 9.5877,
|
| 670 |
+
"eval_samples_per_second": 24.615,
|
| 671 |
+
"step": 25500
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 2.76,
|
| 675 |
+
"learning_rate": 6e-05,
|
| 676 |
+
"loss": 3.9163,
|
| 677 |
+
"step": 26000
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 2.76,
|
| 681 |
+
"eval_loss": 3.7216522693634033,
|
| 682 |
+
"eval_runtime": 9.5913,
|
| 683 |
+
"eval_samples_per_second": 24.606,
|
| 684 |
+
"step": 26000
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 2.81,
|
| 688 |
+
"learning_rate": 6e-05,
|
| 689 |
+
"loss": 3.9174,
|
| 690 |
+
"step": 26500
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 2.81,
|
| 694 |
+
"eval_loss": 3.7126359939575195,
|
| 695 |
+
"eval_runtime": 9.5904,
|
| 696 |
+
"eval_samples_per_second": 24.608,
|
| 697 |
+
"step": 26500
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 2.86,
|
| 701 |
+
"learning_rate": 6e-05,
|
| 702 |
+
"loss": 3.9028,
|
| 703 |
+
"step": 27000
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 2.86,
|
| 707 |
+
"eval_loss": 3.7034430503845215,
|
| 708 |
+
"eval_runtime": 9.599,
|
| 709 |
+
"eval_samples_per_second": 24.586,
|
| 710 |
+
"step": 27000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 2.92,
|
| 714 |
+
"learning_rate": 6e-05,
|
| 715 |
+
"loss": 3.9076,
|
| 716 |
+
"step": 27500
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 2.92,
|
| 720 |
+
"eval_loss": 3.7010114192962646,
|
| 721 |
+
"eval_runtime": 9.5974,
|
| 722 |
+
"eval_samples_per_second": 24.59,
|
| 723 |
+
"step": 27500
|
| 724 |
+
}
|
| 725 |
+
],
|
| 726 |
+
"max_steps": 94290,
|
| 727 |
+
"num_train_epochs": 10,
|
| 728 |
+
"total_flos": 3969647640576000.0,
|
| 729 |
+
"trial_name": null,
|
| 730 |
+
"trial_params": null
|
| 731 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0adc9c91656feb712b613d0af00eee2e222a6b2533e79be85530642f0107a27f
|
| 3 |
+
size 2479
|