Training in progress, step 20000, checkpoint
Browse files- checkpoint-20000/config.json +31 -0
- checkpoint-20000/generation_config.json +6 -0
- checkpoint-20000/model.safetensors +3 -0
- checkpoint-20000/optimizer.pt +3 -0
- checkpoint-20000/rng_state.pth +3 -0
- checkpoint-20000/scheduler.pt +3 -0
- checkpoint-20000/special_tokens_map.json +1 -0
- checkpoint-20000/tokenizer.json +0 -0
- checkpoint-20000/tokenizer_config.json +0 -0
- checkpoint-20000/trainer_state.json +3013 -0
- checkpoint-20000/training_args.bin +3 -0
checkpoint-20000/config.json
ADDED
|
@@ -0,0 +1,31 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"activation_function": "gelu_new",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"GPT2LMHeadModel"
|
| 5 |
+
],
|
| 6 |
+
"attn_pdrop": 0.1,
|
| 7 |
+
"bos_token_id": 50256,
|
| 8 |
+
"embd_pdrop": 0.1,
|
| 9 |
+
"eos_token_id": 50256,
|
| 10 |
+
"initializer_range": 0.02,
|
| 11 |
+
"layer_norm_epsilon": 1e-05,
|
| 12 |
+
"model_type": "gpt2",
|
| 13 |
+
"n_embd": 768,
|
| 14 |
+
"n_head": 12,
|
| 15 |
+
"n_inner": null,
|
| 16 |
+
"n_layer": 12,
|
| 17 |
+
"n_positions": 1024,
|
| 18 |
+
"reorder_and_upcast_attn": false,
|
| 19 |
+
"resid_pdrop": 0.1,
|
| 20 |
+
"scale_attn_by_inverse_layer_idx": false,
|
| 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 |
+
"torch_dtype": "float32",
|
| 28 |
+
"transformers_version": "4.47.0.dev0",
|
| 29 |
+
"use_cache": true,
|
| 30 |
+
"vocab_size": 52000
|
| 31 |
+
}
|
checkpoint-20000/generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 50256,
|
| 4 |
+
"eos_token_id": 50256,
|
| 5 |
+
"transformers_version": "4.47.0.dev0"
|
| 6 |
+
}
|
checkpoint-20000/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:64e29bf0f6db6bf47fb5a720e2b3b07f95efbc47e7436a23c9b559daa81e2100
|
| 3 |
+
size 503128704
|
checkpoint-20000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d3719bed72cfa0a22a121f77a640145cdde9ca1035ef015ce11ad44065bd6eb5
|
| 3 |
+
size 1006351290
|
checkpoint-20000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:909d185b63a506ab3e9eeca3c20a4431427a225cc61d94c40272036ac3f89f08
|
| 3 |
+
size 14244
|
checkpoint-20000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3a5f17012f60b7aaef69b1774fc343935b94271b1567050a7a78b6b97f360210
|
| 3 |
+
size 1064
|
checkpoint-20000/special_tokens_map.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{}
|
checkpoint-20000/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-20000/tokenizer_config.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-20000/trainer_state.json
ADDED
|
@@ -0,0 +1,3013 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 3.5728728771209717,
|
| 3 |
+
"best_model_checkpoint": "/scratch/cl5625/exceptions/models/100M__634/checkpoint-20000",
|
| 4 |
+
"epoch": 2.1563342318059298,
|
| 5 |
+
"eval_steps": 1000,
|
| 6 |
+
"global_step": 20000,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.005390835579514825,
|
| 13 |
+
"grad_norm": 1.412644624710083,
|
| 14 |
+
"learning_rate": 0.0003,
|
| 15 |
+
"loss": 8.6259,
|
| 16 |
+
"step": 50
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.01078167115902965,
|
| 20 |
+
"grad_norm": 3.4297802448272705,
|
| 21 |
+
"learning_rate": 0.0006,
|
| 22 |
+
"loss": 6.925,
|
| 23 |
+
"step": 100
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.016172506738544475,
|
| 27 |
+
"grad_norm": 1.0797241926193237,
|
| 28 |
+
"learning_rate": 0.0005996762007555315,
|
| 29 |
+
"loss": 6.4904,
|
| 30 |
+
"step": 150
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.0215633423180593,
|
| 34 |
+
"grad_norm": 1.8681477308273315,
|
| 35 |
+
"learning_rate": 0.000599352401511063,
|
| 36 |
+
"loss": 6.2514,
|
| 37 |
+
"step": 200
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.026954177897574125,
|
| 41 |
+
"grad_norm": 1.0239183902740479,
|
| 42 |
+
"learning_rate": 0.0005990286022665946,
|
| 43 |
+
"loss": 6.0911,
|
| 44 |
+
"step": 250
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.03234501347708895,
|
| 48 |
+
"grad_norm": 1.5482831001281738,
|
| 49 |
+
"learning_rate": 0.0005987048030221263,
|
| 50 |
+
"loss": 5.9845,
|
| 51 |
+
"step": 300
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.03773584905660377,
|
| 55 |
+
"grad_norm": 1.6334092617034912,
|
| 56 |
+
"learning_rate": 0.0005983810037776578,
|
| 57 |
+
"loss": 5.875,
|
| 58 |
+
"step": 350
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.0431266846361186,
|
| 62 |
+
"grad_norm": 1.5844093561172485,
|
| 63 |
+
"learning_rate": 0.0005980572045331894,
|
| 64 |
+
"loss": 5.7818,
|
| 65 |
+
"step": 400
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.04851752021563342,
|
| 69 |
+
"grad_norm": 1.2689578533172607,
|
| 70 |
+
"learning_rate": 0.0005977334052887209,
|
| 71 |
+
"loss": 5.712,
|
| 72 |
+
"step": 450
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.05390835579514825,
|
| 76 |
+
"grad_norm": 1.3288975954055786,
|
| 77 |
+
"learning_rate": 0.0005974096060442526,
|
| 78 |
+
"loss": 5.6535,
|
| 79 |
+
"step": 500
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.05929919137466307,
|
| 83 |
+
"grad_norm": 2.041428804397583,
|
| 84 |
+
"learning_rate": 0.0005970858067997841,
|
| 85 |
+
"loss": 5.5725,
|
| 86 |
+
"step": 550
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.0646900269541779,
|
| 90 |
+
"grad_norm": 1.4345282316207886,
|
| 91 |
+
"learning_rate": 0.0005967620075553157,
|
| 92 |
+
"loss": 5.5081,
|
| 93 |
+
"step": 600
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.07008086253369272,
|
| 97 |
+
"grad_norm": 1.1454274654388428,
|
| 98 |
+
"learning_rate": 0.0005964382083108472,
|
| 99 |
+
"loss": 5.4223,
|
| 100 |
+
"step": 650
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.07547169811320754,
|
| 104 |
+
"grad_norm": 1.3169114589691162,
|
| 105 |
+
"learning_rate": 0.0005961144090663788,
|
| 106 |
+
"loss": 5.3711,
|
| 107 |
+
"step": 700
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.08086253369272237,
|
| 111 |
+
"grad_norm": 1.3919389247894287,
|
| 112 |
+
"learning_rate": 0.0005957906098219104,
|
| 113 |
+
"loss": 5.3003,
|
| 114 |
+
"step": 750
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.0862533692722372,
|
| 118 |
+
"grad_norm": 1.0709481239318848,
|
| 119 |
+
"learning_rate": 0.0005954668105774419,
|
| 120 |
+
"loss": 5.2637,
|
| 121 |
+
"step": 800
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.09164420485175202,
|
| 125 |
+
"grad_norm": 0.8944941759109497,
|
| 126 |
+
"learning_rate": 0.0005951430113329735,
|
| 127 |
+
"loss": 5.1967,
|
| 128 |
+
"step": 850
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.09703504043126684,
|
| 132 |
+
"grad_norm": 1.1984753608703613,
|
| 133 |
+
"learning_rate": 0.0005948192120885051,
|
| 134 |
+
"loss": 5.1709,
|
| 135 |
+
"step": 900
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.10242587601078167,
|
| 139 |
+
"grad_norm": 1.0900869369506836,
|
| 140 |
+
"learning_rate": 0.0005944954128440366,
|
| 141 |
+
"loss": 5.1332,
|
| 142 |
+
"step": 950
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.1078167115902965,
|
| 146 |
+
"grad_norm": 1.4634649753570557,
|
| 147 |
+
"learning_rate": 0.0005941716135995682,
|
| 148 |
+
"loss": 5.0778,
|
| 149 |
+
"step": 1000
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.1078167115902965,
|
| 153 |
+
"eval_accuracy": 0.22697957684263617,
|
| 154 |
+
"eval_loss": 5.024423599243164,
|
| 155 |
+
"eval_runtime": 183.2168,
|
| 156 |
+
"eval_samples_per_second": 98.304,
|
| 157 |
+
"eval_steps_per_second": 6.146,
|
| 158 |
+
"step": 1000
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.11320754716981132,
|
| 162 |
+
"grad_norm": 1.0724071264266968,
|
| 163 |
+
"learning_rate": 0.0005938478143550997,
|
| 164 |
+
"loss": 5.039,
|
| 165 |
+
"step": 1050
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 0.11859838274932614,
|
| 169 |
+
"grad_norm": 0.7434335947036743,
|
| 170 |
+
"learning_rate": 0.0005935240151106314,
|
| 171 |
+
"loss": 5.0145,
|
| 172 |
+
"step": 1100
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 0.12398921832884097,
|
| 176 |
+
"grad_norm": 1.1407990455627441,
|
| 177 |
+
"learning_rate": 0.0005932002158661629,
|
| 178 |
+
"loss": 4.9884,
|
| 179 |
+
"step": 1150
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 0.1293800539083558,
|
| 183 |
+
"grad_norm": 1.3579697608947754,
|
| 184 |
+
"learning_rate": 0.0005928764166216945,
|
| 185 |
+
"loss": 4.9232,
|
| 186 |
+
"step": 1200
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 0.1347708894878706,
|
| 190 |
+
"grad_norm": 1.170091986656189,
|
| 191 |
+
"learning_rate": 0.000592552617377226,
|
| 192 |
+
"loss": 4.9022,
|
| 193 |
+
"step": 1250
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 0.14016172506738545,
|
| 197 |
+
"grad_norm": 1.056681513786316,
|
| 198 |
+
"learning_rate": 0.0005922288181327577,
|
| 199 |
+
"loss": 4.8796,
|
| 200 |
+
"step": 1300
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.14555256064690028,
|
| 204 |
+
"grad_norm": 0.8261628746986389,
|
| 205 |
+
"learning_rate": 0.0005919050188882893,
|
| 206 |
+
"loss": 4.8425,
|
| 207 |
+
"step": 1350
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 0.1509433962264151,
|
| 211 |
+
"grad_norm": 0.7253502011299133,
|
| 212 |
+
"learning_rate": 0.0005915812196438207,
|
| 213 |
+
"loss": 4.836,
|
| 214 |
+
"step": 1400
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 0.15633423180592992,
|
| 218 |
+
"grad_norm": 1.0881081819534302,
|
| 219 |
+
"learning_rate": 0.0005912574203993524,
|
| 220 |
+
"loss": 4.8436,
|
| 221 |
+
"step": 1450
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 0.16172506738544473,
|
| 225 |
+
"grad_norm": 0.8581550717353821,
|
| 226 |
+
"learning_rate": 0.0005909336211548839,
|
| 227 |
+
"loss": 4.8009,
|
| 228 |
+
"step": 1500
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 0.16711590296495957,
|
| 232 |
+
"grad_norm": 0.9875918030738831,
|
| 233 |
+
"learning_rate": 0.0005906098219104155,
|
| 234 |
+
"loss": 4.7541,
|
| 235 |
+
"step": 1550
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 0.1725067385444744,
|
| 239 |
+
"grad_norm": 0.8188138008117676,
|
| 240 |
+
"learning_rate": 0.000590286022665947,
|
| 241 |
+
"loss": 4.744,
|
| 242 |
+
"step": 1600
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 0.1778975741239892,
|
| 246 |
+
"grad_norm": 0.8326888084411621,
|
| 247 |
+
"learning_rate": 0.0005899622234214787,
|
| 248 |
+
"loss": 4.7177,
|
| 249 |
+
"step": 1650
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 0.18328840970350405,
|
| 253 |
+
"grad_norm": 1.0679171085357666,
|
| 254 |
+
"learning_rate": 0.0005896384241770102,
|
| 255 |
+
"loss": 4.6933,
|
| 256 |
+
"step": 1700
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 0.18867924528301888,
|
| 260 |
+
"grad_norm": 0.8685047626495361,
|
| 261 |
+
"learning_rate": 0.0005893146249325418,
|
| 262 |
+
"loss": 4.7025,
|
| 263 |
+
"step": 1750
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 0.1940700808625337,
|
| 267 |
+
"grad_norm": 0.8494108319282532,
|
| 268 |
+
"learning_rate": 0.0005889908256880733,
|
| 269 |
+
"loss": 4.6703,
|
| 270 |
+
"step": 1800
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 0.19946091644204852,
|
| 274 |
+
"grad_norm": 1.175525426864624,
|
| 275 |
+
"learning_rate": 0.0005886670264436049,
|
| 276 |
+
"loss": 4.6366,
|
| 277 |
+
"step": 1850
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 0.20485175202156333,
|
| 281 |
+
"grad_norm": 1.0531870126724243,
|
| 282 |
+
"learning_rate": 0.0005883432271991365,
|
| 283 |
+
"loss": 4.6263,
|
| 284 |
+
"step": 1900
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.21024258760107817,
|
| 288 |
+
"grad_norm": 0.9467228651046753,
|
| 289 |
+
"learning_rate": 0.0005880194279546681,
|
| 290 |
+
"loss": 4.6254,
|
| 291 |
+
"step": 1950
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 0.215633423180593,
|
| 295 |
+
"grad_norm": 0.9411425590515137,
|
| 296 |
+
"learning_rate": 0.0005876956287101996,
|
| 297 |
+
"loss": 4.5738,
|
| 298 |
+
"step": 2000
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 0.215633423180593,
|
| 302 |
+
"eval_accuracy": 0.271918085284042,
|
| 303 |
+
"eval_loss": 4.5008015632629395,
|
| 304 |
+
"eval_runtime": 181.4309,
|
| 305 |
+
"eval_samples_per_second": 99.272,
|
| 306 |
+
"eval_steps_per_second": 6.206,
|
| 307 |
+
"step": 2000
|
| 308 |
+
},
|
| 309 |
+
{
|
| 310 |
+
"epoch": 0.2210242587601078,
|
| 311 |
+
"grad_norm": 0.920215368270874,
|
| 312 |
+
"learning_rate": 0.0005873718294657312,
|
| 313 |
+
"loss": 4.5559,
|
| 314 |
+
"step": 2050
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.22641509433962265,
|
| 318 |
+
"grad_norm": 0.7012114524841309,
|
| 319 |
+
"learning_rate": 0.0005870480302212628,
|
| 320 |
+
"loss": 4.5445,
|
| 321 |
+
"step": 2100
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 0.23180592991913745,
|
| 325 |
+
"grad_norm": 0.9538075923919678,
|
| 326 |
+
"learning_rate": 0.0005867242309767943,
|
| 327 |
+
"loss": 4.5259,
|
| 328 |
+
"step": 2150
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 0.2371967654986523,
|
| 332 |
+
"grad_norm": 0.7847324013710022,
|
| 333 |
+
"learning_rate": 0.0005864004317323259,
|
| 334 |
+
"loss": 4.4889,
|
| 335 |
+
"step": 2200
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"epoch": 0.24258760107816713,
|
| 339 |
+
"grad_norm": 1.0440484285354614,
|
| 340 |
+
"learning_rate": 0.0005860766324878575,
|
| 341 |
+
"loss": 4.504,
|
| 342 |
+
"step": 2250
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 0.24797843665768193,
|
| 346 |
+
"grad_norm": 0.782093346118927,
|
| 347 |
+
"learning_rate": 0.000585752833243389,
|
| 348 |
+
"loss": 4.4768,
|
| 349 |
+
"step": 2300
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"epoch": 0.25336927223719674,
|
| 353 |
+
"grad_norm": 0.9644035696983337,
|
| 354 |
+
"learning_rate": 0.0005854290339989206,
|
| 355 |
+
"loss": 4.4522,
|
| 356 |
+
"step": 2350
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.2587601078167116,
|
| 360 |
+
"grad_norm": 1.0412849187850952,
|
| 361 |
+
"learning_rate": 0.0005851052347544521,
|
| 362 |
+
"loss": 4.4489,
|
| 363 |
+
"step": 2400
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 0.2641509433962264,
|
| 367 |
+
"grad_norm": 0.7987921237945557,
|
| 368 |
+
"learning_rate": 0.0005847814355099838,
|
| 369 |
+
"loss": 4.4304,
|
| 370 |
+
"step": 2450
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 0.2695417789757412,
|
| 374 |
+
"grad_norm": 0.8646295070648193,
|
| 375 |
+
"learning_rate": 0.0005844576362655154,
|
| 376 |
+
"loss": 4.4196,
|
| 377 |
+
"step": 2500
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 0.2749326145552561,
|
| 381 |
+
"grad_norm": 1.0314178466796875,
|
| 382 |
+
"learning_rate": 0.0005841338370210469,
|
| 383 |
+
"loss": 4.4017,
|
| 384 |
+
"step": 2550
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 0.2803234501347709,
|
| 388 |
+
"grad_norm": 0.9287024140357971,
|
| 389 |
+
"learning_rate": 0.0005838100377765785,
|
| 390 |
+
"loss": 4.3729,
|
| 391 |
+
"step": 2600
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 0.2857142857142857,
|
| 395 |
+
"grad_norm": 0.8492061495780945,
|
| 396 |
+
"learning_rate": 0.0005834862385321101,
|
| 397 |
+
"loss": 4.37,
|
| 398 |
+
"step": 2650
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 0.29110512129380056,
|
| 402 |
+
"grad_norm": 0.7161704301834106,
|
| 403 |
+
"learning_rate": 0.0005831624392876417,
|
| 404 |
+
"loss": 4.3608,
|
| 405 |
+
"step": 2700
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 0.29649595687331537,
|
| 409 |
+
"grad_norm": 1.0816575288772583,
|
| 410 |
+
"learning_rate": 0.0005828386400431731,
|
| 411 |
+
"loss": 4.3705,
|
| 412 |
+
"step": 2750
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 0.3018867924528302,
|
| 416 |
+
"grad_norm": 0.8127713799476624,
|
| 417 |
+
"learning_rate": 0.0005825148407987048,
|
| 418 |
+
"loss": 4.3681,
|
| 419 |
+
"step": 2800
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 0.30727762803234504,
|
| 423 |
+
"grad_norm": 0.9167231321334839,
|
| 424 |
+
"learning_rate": 0.0005821910415542363,
|
| 425 |
+
"loss": 4.33,
|
| 426 |
+
"step": 2850
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 0.31266846361185985,
|
| 430 |
+
"grad_norm": 0.8297504782676697,
|
| 431 |
+
"learning_rate": 0.0005818672423097679,
|
| 432 |
+
"loss": 4.3137,
|
| 433 |
+
"step": 2900
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 0.31805929919137466,
|
| 437 |
+
"grad_norm": 0.9186894297599792,
|
| 438 |
+
"learning_rate": 0.0005815434430652994,
|
| 439 |
+
"loss": 4.3441,
|
| 440 |
+
"step": 2950
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 0.32345013477088946,
|
| 444 |
+
"grad_norm": 0.7359763979911804,
|
| 445 |
+
"learning_rate": 0.0005812196438208311,
|
| 446 |
+
"loss": 4.309,
|
| 447 |
+
"step": 3000
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 0.32345013477088946,
|
| 451 |
+
"eval_accuracy": 0.29948495271589304,
|
| 452 |
+
"eval_loss": 4.228354454040527,
|
| 453 |
+
"eval_runtime": 183.1945,
|
| 454 |
+
"eval_samples_per_second": 98.316,
|
| 455 |
+
"eval_steps_per_second": 6.146,
|
| 456 |
+
"step": 3000
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 0.3288409703504043,
|
| 460 |
+
"grad_norm": 0.7310630679130554,
|
| 461 |
+
"learning_rate": 0.0005808958445763626,
|
| 462 |
+
"loss": 4.3024,
|
| 463 |
+
"step": 3050
|
| 464 |
+
},
|
| 465 |
+
{
|
| 466 |
+
"epoch": 0.33423180592991913,
|
| 467 |
+
"grad_norm": 0.8260939717292786,
|
| 468 |
+
"learning_rate": 0.0005805720453318942,
|
| 469 |
+
"loss": 4.2831,
|
| 470 |
+
"step": 3100
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 0.33962264150943394,
|
| 474 |
+
"grad_norm": 0.7478753924369812,
|
| 475 |
+
"learning_rate": 0.0005802482460874257,
|
| 476 |
+
"loss": 4.2722,
|
| 477 |
+
"step": 3150
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 0.3450134770889488,
|
| 481 |
+
"grad_norm": 0.8138642907142639,
|
| 482 |
+
"learning_rate": 0.0005799244468429573,
|
| 483 |
+
"loss": 4.2816,
|
| 484 |
+
"step": 3200
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 0.3504043126684636,
|
| 488 |
+
"grad_norm": 0.7284408211708069,
|
| 489 |
+
"learning_rate": 0.0005796006475984889,
|
| 490 |
+
"loss": 4.2495,
|
| 491 |
+
"step": 3250
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 0.3557951482479784,
|
| 495 |
+
"grad_norm": 0.6755571365356445,
|
| 496 |
+
"learning_rate": 0.0005792768483540205,
|
| 497 |
+
"loss": 4.2595,
|
| 498 |
+
"step": 3300
|
| 499 |
+
},
|
| 500 |
+
{
|
| 501 |
+
"epoch": 0.3611859838274933,
|
| 502 |
+
"grad_norm": 0.9542193412780762,
|
| 503 |
+
"learning_rate": 0.000578953049109552,
|
| 504 |
+
"loss": 4.2538,
|
| 505 |
+
"step": 3350
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 0.3665768194070081,
|
| 509 |
+
"grad_norm": 0.6383764743804932,
|
| 510 |
+
"learning_rate": 0.0005786292498650836,
|
| 511 |
+
"loss": 4.2319,
|
| 512 |
+
"step": 3400
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 0.3719676549865229,
|
| 516 |
+
"grad_norm": 0.8126682639122009,
|
| 517 |
+
"learning_rate": 0.0005783054506206152,
|
| 518 |
+
"loss": 4.2272,
|
| 519 |
+
"step": 3450
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 0.37735849056603776,
|
| 523 |
+
"grad_norm": 0.7132463455200195,
|
| 524 |
+
"learning_rate": 0.0005779816513761467,
|
| 525 |
+
"loss": 4.2186,
|
| 526 |
+
"step": 3500
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 0.38274932614555257,
|
| 530 |
+
"grad_norm": 0.818424642086029,
|
| 531 |
+
"learning_rate": 0.0005776578521316782,
|
| 532 |
+
"loss": 4.1982,
|
| 533 |
+
"step": 3550
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 0.3881401617250674,
|
| 537 |
+
"grad_norm": 0.6916500926017761,
|
| 538 |
+
"learning_rate": 0.0005773340528872099,
|
| 539 |
+
"loss": 4.2057,
|
| 540 |
+
"step": 3600
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 0.3935309973045822,
|
| 544 |
+
"grad_norm": 0.6858584880828857,
|
| 545 |
+
"learning_rate": 0.0005770102536427414,
|
| 546 |
+
"loss": 4.2202,
|
| 547 |
+
"step": 3650
|
| 548 |
+
},
|
| 549 |
+
{
|
| 550 |
+
"epoch": 0.39892183288409705,
|
| 551 |
+
"grad_norm": 0.7574723362922668,
|
| 552 |
+
"learning_rate": 0.000576686454398273,
|
| 553 |
+
"loss": 4.2073,
|
| 554 |
+
"step": 3700
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 0.40431266846361186,
|
| 558 |
+
"grad_norm": 0.7107515931129456,
|
| 559 |
+
"learning_rate": 0.0005763626551538045,
|
| 560 |
+
"loss": 4.1939,
|
| 561 |
+
"step": 3750
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 0.40970350404312667,
|
| 565 |
+
"grad_norm": 0.7424683570861816,
|
| 566 |
+
"learning_rate": 0.0005760388559093362,
|
| 567 |
+
"loss": 4.1827,
|
| 568 |
+
"step": 3800
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 0.41509433962264153,
|
| 572 |
+
"grad_norm": 0.6541684865951538,
|
| 573 |
+
"learning_rate": 0.0005757150566648678,
|
| 574 |
+
"loss": 4.1782,
|
| 575 |
+
"step": 3850
|
| 576 |
+
},
|
| 577 |
+
{
|
| 578 |
+
"epoch": 0.42048517520215634,
|
| 579 |
+
"grad_norm": 0.797566294670105,
|
| 580 |
+
"learning_rate": 0.0005753912574203993,
|
| 581 |
+
"loss": 4.17,
|
| 582 |
+
"step": 3900
|
| 583 |
+
},
|
| 584 |
+
{
|
| 585 |
+
"epoch": 0.42587601078167114,
|
| 586 |
+
"grad_norm": 0.5999880433082581,
|
| 587 |
+
"learning_rate": 0.0005750674581759309,
|
| 588 |
+
"loss": 4.1651,
|
| 589 |
+
"step": 3950
|
| 590 |
+
},
|
| 591 |
+
{
|
| 592 |
+
"epoch": 0.431266846361186,
|
| 593 |
+
"grad_norm": 0.8264360427856445,
|
| 594 |
+
"learning_rate": 0.0005747436589314624,
|
| 595 |
+
"loss": 4.1524,
|
| 596 |
+
"step": 4000
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 0.431266846361186,
|
| 600 |
+
"eval_accuracy": 0.312240258480874,
|
| 601 |
+
"eval_loss": 4.087605953216553,
|
| 602 |
+
"eval_runtime": 183.2024,
|
| 603 |
+
"eval_samples_per_second": 98.312,
|
| 604 |
+
"eval_steps_per_second": 6.146,
|
| 605 |
+
"step": 4000
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.4366576819407008,
|
| 609 |
+
"grad_norm": 0.7332233786582947,
|
| 610 |
+
"learning_rate": 0.0005744198596869941,
|
| 611 |
+
"loss": 4.1654,
|
| 612 |
+
"step": 4050
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.4420485175202156,
|
| 616 |
+
"grad_norm": 0.5775137543678284,
|
| 617 |
+
"learning_rate": 0.0005740960604425255,
|
| 618 |
+
"loss": 4.158,
|
| 619 |
+
"step": 4100
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.4474393530997305,
|
| 623 |
+
"grad_norm": 0.8009674549102783,
|
| 624 |
+
"learning_rate": 0.0005737722611980572,
|
| 625 |
+
"loss": 4.1392,
|
| 626 |
+
"step": 4150
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.4528301886792453,
|
| 630 |
+
"grad_norm": 0.6772514581680298,
|
| 631 |
+
"learning_rate": 0.0005734484619535887,
|
| 632 |
+
"loss": 4.1186,
|
| 633 |
+
"step": 4200
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.4582210242587601,
|
| 637 |
+
"grad_norm": 0.6068300604820251,
|
| 638 |
+
"learning_rate": 0.0005731246627091203,
|
| 639 |
+
"loss": 4.1518,
|
| 640 |
+
"step": 4250
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.4636118598382749,
|
| 644 |
+
"grad_norm": 0.6603842973709106,
|
| 645 |
+
"learning_rate": 0.0005728008634646518,
|
| 646 |
+
"loss": 4.1228,
|
| 647 |
+
"step": 4300
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.46900269541778977,
|
| 651 |
+
"grad_norm": 0.6870344877243042,
|
| 652 |
+
"learning_rate": 0.0005724770642201835,
|
| 653 |
+
"loss": 4.1306,
|
| 654 |
+
"step": 4350
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.4743935309973046,
|
| 658 |
+
"grad_norm": 0.8701305985450745,
|
| 659 |
+
"learning_rate": 0.000572153264975715,
|
| 660 |
+
"loss": 4.1236,
|
| 661 |
+
"step": 4400
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.4797843665768194,
|
| 665 |
+
"grad_norm": 0.646145224571228,
|
| 666 |
+
"learning_rate": 0.0005718294657312466,
|
| 667 |
+
"loss": 4.0979,
|
| 668 |
+
"step": 4450
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.48517520215633425,
|
| 672 |
+
"grad_norm": 0.5847651362419128,
|
| 673 |
+
"learning_rate": 0.0005715056664867781,
|
| 674 |
+
"loss": 4.0943,
|
| 675 |
+
"step": 4500
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.49056603773584906,
|
| 679 |
+
"grad_norm": 0.7485958933830261,
|
| 680 |
+
"learning_rate": 0.0005711818672423097,
|
| 681 |
+
"loss": 4.0872,
|
| 682 |
+
"step": 4550
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.49595687331536387,
|
| 686 |
+
"grad_norm": 0.6113314628601074,
|
| 687 |
+
"learning_rate": 0.0005708580679978413,
|
| 688 |
+
"loss": 4.0971,
|
| 689 |
+
"step": 4600
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.5013477088948787,
|
| 693 |
+
"grad_norm": 0.814666748046875,
|
| 694 |
+
"learning_rate": 0.0005705342687533729,
|
| 695 |
+
"loss": 4.088,
|
| 696 |
+
"step": 4650
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.5067385444743935,
|
| 700 |
+
"grad_norm": 0.6813623905181885,
|
| 701 |
+
"learning_rate": 0.0005702104695089044,
|
| 702 |
+
"loss": 4.0697,
|
| 703 |
+
"step": 4700
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.5121293800539084,
|
| 707 |
+
"grad_norm": 0.7049392461776733,
|
| 708 |
+
"learning_rate": 0.000569886670264436,
|
| 709 |
+
"loss": 4.083,
|
| 710 |
+
"step": 4750
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.5175202156334232,
|
| 714 |
+
"grad_norm": 0.6689280867576599,
|
| 715 |
+
"learning_rate": 0.0005695628710199675,
|
| 716 |
+
"loss": 4.0704,
|
| 717 |
+
"step": 4800
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.522911051212938,
|
| 721 |
+
"grad_norm": 0.5872611403465271,
|
| 722 |
+
"learning_rate": 0.0005692390717754991,
|
| 723 |
+
"loss": 4.0659,
|
| 724 |
+
"step": 4850
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 0.5283018867924528,
|
| 728 |
+
"grad_norm": 0.7064571976661682,
|
| 729 |
+
"learning_rate": 0.0005689152725310306,
|
| 730 |
+
"loss": 4.0586,
|
| 731 |
+
"step": 4900
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.5336927223719676,
|
| 735 |
+
"grad_norm": 0.7021653652191162,
|
| 736 |
+
"learning_rate": 0.0005685914732865623,
|
| 737 |
+
"loss": 4.0619,
|
| 738 |
+
"step": 4950
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.5390835579514824,
|
| 742 |
+
"grad_norm": 0.6203348636627197,
|
| 743 |
+
"learning_rate": 0.0005682676740420939,
|
| 744 |
+
"loss": 4.0714,
|
| 745 |
+
"step": 5000
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.5390835579514824,
|
| 749 |
+
"eval_accuracy": 0.32160700663358477,
|
| 750 |
+
"eval_loss": 3.9901158809661865,
|
| 751 |
+
"eval_runtime": 183.2916,
|
| 752 |
+
"eval_samples_per_second": 98.264,
|
| 753 |
+
"eval_steps_per_second": 6.143,
|
| 754 |
+
"step": 5000
|
| 755 |
+
},
|
| 756 |
+
{
|
| 757 |
+
"epoch": 0.5444743935309974,
|
| 758 |
+
"grad_norm": 0.7498524188995361,
|
| 759 |
+
"learning_rate": 0.0005679438747976254,
|
| 760 |
+
"loss": 4.0504,
|
| 761 |
+
"step": 5050
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 0.5498652291105122,
|
| 765 |
+
"grad_norm": 0.6415616869926453,
|
| 766 |
+
"learning_rate": 0.000567620075553157,
|
| 767 |
+
"loss": 4.045,
|
| 768 |
+
"step": 5100
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 0.555256064690027,
|
| 772 |
+
"grad_norm": 0.68656986951828,
|
| 773 |
+
"learning_rate": 0.0005672962763086886,
|
| 774 |
+
"loss": 4.0469,
|
| 775 |
+
"step": 5150
|
| 776 |
+
},
|
| 777 |
+
{
|
| 778 |
+
"epoch": 0.5606469002695418,
|
| 779 |
+
"grad_norm": 0.6866742968559265,
|
| 780 |
+
"learning_rate": 0.0005669724770642202,
|
| 781 |
+
"loss": 4.0368,
|
| 782 |
+
"step": 5200
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 0.5660377358490566,
|
| 786 |
+
"grad_norm": 0.6404621601104736,
|
| 787 |
+
"learning_rate": 0.0005666486778197517,
|
| 788 |
+
"loss": 4.0323,
|
| 789 |
+
"step": 5250
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 0.5714285714285714,
|
| 793 |
+
"grad_norm": 0.6272525191307068,
|
| 794 |
+
"learning_rate": 0.0005663248785752833,
|
| 795 |
+
"loss": 4.0431,
|
| 796 |
+
"step": 5300
|
| 797 |
+
},
|
| 798 |
+
{
|
| 799 |
+
"epoch": 0.5768194070080862,
|
| 800 |
+
"grad_norm": 0.6219531297683716,
|
| 801 |
+
"learning_rate": 0.0005660010793308148,
|
| 802 |
+
"loss": 4.0283,
|
| 803 |
+
"step": 5350
|
| 804 |
+
},
|
| 805 |
+
{
|
| 806 |
+
"epoch": 0.5822102425876011,
|
| 807 |
+
"grad_norm": 0.5969902276992798,
|
| 808 |
+
"learning_rate": 0.0005656772800863465,
|
| 809 |
+
"loss": 4.0455,
|
| 810 |
+
"step": 5400
|
| 811 |
+
},
|
| 812 |
+
{
|
| 813 |
+
"epoch": 0.5876010781671159,
|
| 814 |
+
"grad_norm": 0.7162837982177734,
|
| 815 |
+
"learning_rate": 0.0005653534808418779,
|
| 816 |
+
"loss": 4.0187,
|
| 817 |
+
"step": 5450
|
| 818 |
+
},
|
| 819 |
+
{
|
| 820 |
+
"epoch": 0.5929919137466307,
|
| 821 |
+
"grad_norm": 0.5840233564376831,
|
| 822 |
+
"learning_rate": 0.0005650296815974096,
|
| 823 |
+
"loss": 4.0264,
|
| 824 |
+
"step": 5500
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 0.5983827493261455,
|
| 828 |
+
"grad_norm": 0.5966022610664368,
|
| 829 |
+
"learning_rate": 0.0005647058823529411,
|
| 830 |
+
"loss": 4.0154,
|
| 831 |
+
"step": 5550
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 0.6037735849056604,
|
| 835 |
+
"grad_norm": 0.5551064610481262,
|
| 836 |
+
"learning_rate": 0.0005643820831084727,
|
| 837 |
+
"loss": 4.0069,
|
| 838 |
+
"step": 5600
|
| 839 |
+
},
|
| 840 |
+
{
|
| 841 |
+
"epoch": 0.6091644204851752,
|
| 842 |
+
"grad_norm": 0.7473669052124023,
|
| 843 |
+
"learning_rate": 0.0005640582838640042,
|
| 844 |
+
"loss": 4.0204,
|
| 845 |
+
"step": 5650
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 0.6145552560646901,
|
| 849 |
+
"grad_norm": 0.6089016795158386,
|
| 850 |
+
"learning_rate": 0.0005637344846195358,
|
| 851 |
+
"loss": 4.0132,
|
| 852 |
+
"step": 5700
|
| 853 |
+
},
|
| 854 |
+
{
|
| 855 |
+
"epoch": 0.6199460916442049,
|
| 856 |
+
"grad_norm": 0.5742376446723938,
|
| 857 |
+
"learning_rate": 0.0005634106853750674,
|
| 858 |
+
"loss": 4.02,
|
| 859 |
+
"step": 5750
|
| 860 |
+
},
|
| 861 |
+
{
|
| 862 |
+
"epoch": 0.6253369272237197,
|
| 863 |
+
"grad_norm": 0.857323169708252,
|
| 864 |
+
"learning_rate": 0.000563086886130599,
|
| 865 |
+
"loss": 3.9884,
|
| 866 |
+
"step": 5800
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 0.6307277628032345,
|
| 870 |
+
"grad_norm": 0.6711622476577759,
|
| 871 |
+
"learning_rate": 0.0005627630868861305,
|
| 872 |
+
"loss": 3.9889,
|
| 873 |
+
"step": 5850
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 0.6361185983827493,
|
| 877 |
+
"grad_norm": 0.6008835434913635,
|
| 878 |
+
"learning_rate": 0.0005624392876416621,
|
| 879 |
+
"loss": 3.9763,
|
| 880 |
+
"step": 5900
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"epoch": 0.6415094339622641,
|
| 884 |
+
"grad_norm": 0.6592042446136475,
|
| 885 |
+
"learning_rate": 0.0005621154883971937,
|
| 886 |
+
"loss": 3.9816,
|
| 887 |
+
"step": 5950
|
| 888 |
+
},
|
| 889 |
+
{
|
| 890 |
+
"epoch": 0.6469002695417789,
|
| 891 |
+
"grad_norm": 0.6811545491218567,
|
| 892 |
+
"learning_rate": 0.0005617916891527253,
|
| 893 |
+
"loss": 3.9631,
|
| 894 |
+
"step": 6000
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 0.6469002695417789,
|
| 898 |
+
"eval_accuracy": 0.328109556010618,
|
| 899 |
+
"eval_loss": 3.9180023670196533,
|
| 900 |
+
"eval_runtime": 183.5563,
|
| 901 |
+
"eval_samples_per_second": 98.122,
|
| 902 |
+
"eval_steps_per_second": 6.134,
|
| 903 |
+
"step": 6000
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 0.6522911051212938,
|
| 907 |
+
"grad_norm": 0.838979959487915,
|
| 908 |
+
"learning_rate": 0.0005614678899082568,
|
| 909 |
+
"loss": 3.9977,
|
| 910 |
+
"step": 6050
|
| 911 |
+
},
|
| 912 |
+
{
|
| 913 |
+
"epoch": 0.6576819407008087,
|
| 914 |
+
"grad_norm": 0.6069373488426208,
|
| 915 |
+
"learning_rate": 0.0005611440906637884,
|
| 916 |
+
"loss": 3.9844,
|
| 917 |
+
"step": 6100
|
| 918 |
+
},
|
| 919 |
+
{
|
| 920 |
+
"epoch": 0.6630727762803235,
|
| 921 |
+
"grad_norm": 0.6294558048248291,
|
| 922 |
+
"learning_rate": 0.00056082029141932,
|
| 923 |
+
"loss": 3.9769,
|
| 924 |
+
"step": 6150
|
| 925 |
+
},
|
| 926 |
+
{
|
| 927 |
+
"epoch": 0.6684636118598383,
|
| 928 |
+
"grad_norm": 0.6527539491653442,
|
| 929 |
+
"learning_rate": 0.0005604964921748515,
|
| 930 |
+
"loss": 3.9875,
|
| 931 |
+
"step": 6200
|
| 932 |
+
},
|
| 933 |
+
{
|
| 934 |
+
"epoch": 0.6738544474393531,
|
| 935 |
+
"grad_norm": 0.6204699873924255,
|
| 936 |
+
"learning_rate": 0.000560172692930383,
|
| 937 |
+
"loss": 3.949,
|
| 938 |
+
"step": 6250
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 0.6792452830188679,
|
| 942 |
+
"grad_norm": 0.6665420532226562,
|
| 943 |
+
"learning_rate": 0.0005598488936859147,
|
| 944 |
+
"loss": 3.96,
|
| 945 |
+
"step": 6300
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 0.6846361185983828,
|
| 949 |
+
"grad_norm": 0.6489077806472778,
|
| 950 |
+
"learning_rate": 0.0005595250944414463,
|
| 951 |
+
"loss": 3.9534,
|
| 952 |
+
"step": 6350
|
| 953 |
+
},
|
| 954 |
+
{
|
| 955 |
+
"epoch": 0.6900269541778976,
|
| 956 |
+
"grad_norm": 0.6623448729515076,
|
| 957 |
+
"learning_rate": 0.0005592012951969778,
|
| 958 |
+
"loss": 3.9454,
|
| 959 |
+
"step": 6400
|
| 960 |
+
},
|
| 961 |
+
{
|
| 962 |
+
"epoch": 0.6954177897574124,
|
| 963 |
+
"grad_norm": 0.7035852670669556,
|
| 964 |
+
"learning_rate": 0.0005588774959525094,
|
| 965 |
+
"loss": 3.9722,
|
| 966 |
+
"step": 6450
|
| 967 |
+
},
|
| 968 |
+
{
|
| 969 |
+
"epoch": 0.7008086253369272,
|
| 970 |
+
"grad_norm": 0.5686920881271362,
|
| 971 |
+
"learning_rate": 0.000558553696708041,
|
| 972 |
+
"loss": 3.9525,
|
| 973 |
+
"step": 6500
|
| 974 |
+
},
|
| 975 |
+
{
|
| 976 |
+
"epoch": 0.706199460916442,
|
| 977 |
+
"grad_norm": 0.5780633091926575,
|
| 978 |
+
"learning_rate": 0.0005582298974635726,
|
| 979 |
+
"loss": 3.9525,
|
| 980 |
+
"step": 6550
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 0.7115902964959568,
|
| 984 |
+
"grad_norm": 0.6254565715789795,
|
| 985 |
+
"learning_rate": 0.0005579060982191041,
|
| 986 |
+
"loss": 3.9494,
|
| 987 |
+
"step": 6600
|
| 988 |
+
},
|
| 989 |
+
{
|
| 990 |
+
"epoch": 0.7169811320754716,
|
| 991 |
+
"grad_norm": 0.6464234590530396,
|
| 992 |
+
"learning_rate": 0.0005575822989746357,
|
| 993 |
+
"loss": 3.9432,
|
| 994 |
+
"step": 6650
|
| 995 |
+
},
|
| 996 |
+
{
|
| 997 |
+
"epoch": 0.7223719676549866,
|
| 998 |
+
"grad_norm": 0.5895872712135315,
|
| 999 |
+
"learning_rate": 0.0005572584997301672,
|
| 1000 |
+
"loss": 3.9378,
|
| 1001 |
+
"step": 6700
|
| 1002 |
+
},
|
| 1003 |
+
{
|
| 1004 |
+
"epoch": 0.7277628032345014,
|
| 1005 |
+
"grad_norm": 0.6661088466644287,
|
| 1006 |
+
"learning_rate": 0.0005569347004856989,
|
| 1007 |
+
"loss": 3.9405,
|
| 1008 |
+
"step": 6750
|
| 1009 |
+
},
|
| 1010 |
+
{
|
| 1011 |
+
"epoch": 0.7331536388140162,
|
| 1012 |
+
"grad_norm": 0.5486308932304382,
|
| 1013 |
+
"learning_rate": 0.0005566109012412303,
|
| 1014 |
+
"loss": 3.9499,
|
| 1015 |
+
"step": 6800
|
| 1016 |
+
},
|
| 1017 |
+
{
|
| 1018 |
+
"epoch": 0.738544474393531,
|
| 1019 |
+
"grad_norm": 0.5206322073936462,
|
| 1020 |
+
"learning_rate": 0.000556287101996762,
|
| 1021 |
+
"loss": 3.9308,
|
| 1022 |
+
"step": 6850
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 0.7439353099730458,
|
| 1026 |
+
"grad_norm": 0.60262531042099,
|
| 1027 |
+
"learning_rate": 0.0005559633027522935,
|
| 1028 |
+
"loss": 3.9279,
|
| 1029 |
+
"step": 6900
|
| 1030 |
+
},
|
| 1031 |
+
{
|
| 1032 |
+
"epoch": 0.7493261455525606,
|
| 1033 |
+
"grad_norm": 0.6978471875190735,
|
| 1034 |
+
"learning_rate": 0.0005556395035078251,
|
| 1035 |
+
"loss": 3.9152,
|
| 1036 |
+
"step": 6950
|
| 1037 |
+
},
|
| 1038 |
+
{
|
| 1039 |
+
"epoch": 0.7547169811320755,
|
| 1040 |
+
"grad_norm": 0.5400425791740417,
|
| 1041 |
+
"learning_rate": 0.0005553157042633566,
|
| 1042 |
+
"loss": 3.9285,
|
| 1043 |
+
"step": 7000
|
| 1044 |
+
},
|
| 1045 |
+
{
|
| 1046 |
+
"epoch": 0.7547169811320755,
|
| 1047 |
+
"eval_accuracy": 0.33344778104183126,
|
| 1048 |
+
"eval_loss": 3.8615846633911133,
|
| 1049 |
+
"eval_runtime": 182.602,
|
| 1050 |
+
"eval_samples_per_second": 98.635,
|
| 1051 |
+
"eval_steps_per_second": 6.166,
|
| 1052 |
+
"step": 7000
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 0.7601078167115903,
|
| 1056 |
+
"grad_norm": 0.5189043283462524,
|
| 1057 |
+
"learning_rate": 0.0005549919050188882,
|
| 1058 |
+
"loss": 3.9302,
|
| 1059 |
+
"step": 7050
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 0.7654986522911051,
|
| 1063 |
+
"grad_norm": 0.5751085877418518,
|
| 1064 |
+
"learning_rate": 0.0005546681057744198,
|
| 1065 |
+
"loss": 3.9332,
|
| 1066 |
+
"step": 7100
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 0.77088948787062,
|
| 1070 |
+
"grad_norm": 0.6791032552719116,
|
| 1071 |
+
"learning_rate": 0.0005543443065299514,
|
| 1072 |
+
"loss": 3.9073,
|
| 1073 |
+
"step": 7150
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 0.7762803234501348,
|
| 1077 |
+
"grad_norm": 0.7200894951820374,
|
| 1078 |
+
"learning_rate": 0.000554020507285483,
|
| 1079 |
+
"loss": 3.9193,
|
| 1080 |
+
"step": 7200
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 0.7816711590296496,
|
| 1084 |
+
"grad_norm": 0.6101612448692322,
|
| 1085 |
+
"learning_rate": 0.0005536967080410145,
|
| 1086 |
+
"loss": 3.919,
|
| 1087 |
+
"step": 7250
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 0.7870619946091644,
|
| 1091 |
+
"grad_norm": 0.5997413396835327,
|
| 1092 |
+
"learning_rate": 0.0005533729087965462,
|
| 1093 |
+
"loss": 3.9079,
|
| 1094 |
+
"step": 7300
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 0.7924528301886793,
|
| 1098 |
+
"grad_norm": 0.5748898386955261,
|
| 1099 |
+
"learning_rate": 0.0005530491095520777,
|
| 1100 |
+
"loss": 3.9162,
|
| 1101 |
+
"step": 7350
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 0.7978436657681941,
|
| 1105 |
+
"grad_norm": 0.6359212398529053,
|
| 1106 |
+
"learning_rate": 0.0005527253103076093,
|
| 1107 |
+
"loss": 3.8807,
|
| 1108 |
+
"step": 7400
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 0.8032345013477089,
|
| 1112 |
+
"grad_norm": 0.6088876724243164,
|
| 1113 |
+
"learning_rate": 0.0005524015110631408,
|
| 1114 |
+
"loss": 3.8902,
|
| 1115 |
+
"step": 7450
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 0.8086253369272237,
|
| 1119 |
+
"grad_norm": 0.6242630481719971,
|
| 1120 |
+
"learning_rate": 0.0005520777118186724,
|
| 1121 |
+
"loss": 3.9035,
|
| 1122 |
+
"step": 7500
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 0.8140161725067385,
|
| 1126 |
+
"grad_norm": 0.622336208820343,
|
| 1127 |
+
"learning_rate": 0.0005517539125742039,
|
| 1128 |
+
"loss": 3.8986,
|
| 1129 |
+
"step": 7550
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 0.8194070080862533,
|
| 1133 |
+
"grad_norm": 0.5424439311027527,
|
| 1134 |
+
"learning_rate": 0.0005514301133297355,
|
| 1135 |
+
"loss": 3.9166,
|
| 1136 |
+
"step": 7600
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 0.8247978436657682,
|
| 1140 |
+
"grad_norm": 0.5808700323104858,
|
| 1141 |
+
"learning_rate": 0.0005511063140852671,
|
| 1142 |
+
"loss": 3.9005,
|
| 1143 |
+
"step": 7650
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 0.8301886792452831,
|
| 1147 |
+
"grad_norm": 0.5555144548416138,
|
| 1148 |
+
"learning_rate": 0.0005507825148407987,
|
| 1149 |
+
"loss": 3.8722,
|
| 1150 |
+
"step": 7700
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 0.8355795148247979,
|
| 1154 |
+
"grad_norm": 0.7033872604370117,
|
| 1155 |
+
"learning_rate": 0.0005504587155963302,
|
| 1156 |
+
"loss": 3.8852,
|
| 1157 |
+
"step": 7750
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 0.8409703504043127,
|
| 1161 |
+
"grad_norm": 0.5577759742736816,
|
| 1162 |
+
"learning_rate": 0.0005501349163518618,
|
| 1163 |
+
"loss": 3.9008,
|
| 1164 |
+
"step": 7800
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 0.8463611859838275,
|
| 1168 |
+
"grad_norm": Infinity,
|
| 1169 |
+
"learning_rate": 0.0005498175930922827,
|
| 1170 |
+
"loss": 3.888,
|
| 1171 |
+
"step": 7850
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 0.8517520215633423,
|
| 1175 |
+
"grad_norm": 0.6767401099205017,
|
| 1176 |
+
"learning_rate": 0.0005494937938478143,
|
| 1177 |
+
"loss": 3.8863,
|
| 1178 |
+
"step": 7900
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 0.8571428571428571,
|
| 1182 |
+
"grad_norm": 0.5402886271476746,
|
| 1183 |
+
"learning_rate": 0.0005491699946033459,
|
| 1184 |
+
"loss": 3.8885,
|
| 1185 |
+
"step": 7950
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 0.862533692722372,
|
| 1189 |
+
"grad_norm": 0.5724102258682251,
|
| 1190 |
+
"learning_rate": 0.0005488461953588775,
|
| 1191 |
+
"loss": 3.8676,
|
| 1192 |
+
"step": 8000
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 0.862533692722372,
|
| 1196 |
+
"eval_accuracy": 0.3375052058314874,
|
| 1197 |
+
"eval_loss": 3.816683292388916,
|
| 1198 |
+
"eval_runtime": 181.6977,
|
| 1199 |
+
"eval_samples_per_second": 99.126,
|
| 1200 |
+
"eval_steps_per_second": 6.197,
|
| 1201 |
+
"step": 8000
|
| 1202 |
+
},
|
| 1203 |
+
{
|
| 1204 |
+
"epoch": 0.8679245283018868,
|
| 1205 |
+
"grad_norm": 0.5436832308769226,
|
| 1206 |
+
"learning_rate": 0.000548522396114409,
|
| 1207 |
+
"loss": 3.8838,
|
| 1208 |
+
"step": 8050
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 0.8733153638814016,
|
| 1212 |
+
"grad_norm": 0.6710783243179321,
|
| 1213 |
+
"learning_rate": 0.0005481985968699406,
|
| 1214 |
+
"loss": 3.8623,
|
| 1215 |
+
"step": 8100
|
| 1216 |
+
},
|
| 1217 |
+
{
|
| 1218 |
+
"epoch": 0.8787061994609164,
|
| 1219 |
+
"grad_norm": 0.5671558380126953,
|
| 1220 |
+
"learning_rate": 0.0005478747976254721,
|
| 1221 |
+
"loss": 3.8746,
|
| 1222 |
+
"step": 8150
|
| 1223 |
+
},
|
| 1224 |
+
{
|
| 1225 |
+
"epoch": 0.8840970350404312,
|
| 1226 |
+
"grad_norm": 0.6092495322227478,
|
| 1227 |
+
"learning_rate": 0.0005475509983810037,
|
| 1228 |
+
"loss": 3.8635,
|
| 1229 |
+
"step": 8200
|
| 1230 |
+
},
|
| 1231 |
+
{
|
| 1232 |
+
"epoch": 0.889487870619946,
|
| 1233 |
+
"grad_norm": 0.6280654072761536,
|
| 1234 |
+
"learning_rate": 0.0005472271991365352,
|
| 1235 |
+
"loss": 3.8772,
|
| 1236 |
+
"step": 8250
|
| 1237 |
+
},
|
| 1238 |
+
{
|
| 1239 |
+
"epoch": 0.894878706199461,
|
| 1240 |
+
"grad_norm": 0.5402874946594238,
|
| 1241 |
+
"learning_rate": 0.0005469033998920669,
|
| 1242 |
+
"loss": 3.8525,
|
| 1243 |
+
"step": 8300
|
| 1244 |
+
},
|
| 1245 |
+
{
|
| 1246 |
+
"epoch": 0.9002695417789758,
|
| 1247 |
+
"grad_norm": 0.5597426295280457,
|
| 1248 |
+
"learning_rate": 0.0005465796006475984,
|
| 1249 |
+
"loss": 3.8685,
|
| 1250 |
+
"step": 8350
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 0.9056603773584906,
|
| 1254 |
+
"grad_norm": 0.5542333126068115,
|
| 1255 |
+
"learning_rate": 0.00054625580140313,
|
| 1256 |
+
"loss": 3.8677,
|
| 1257 |
+
"step": 8400
|
| 1258 |
+
},
|
| 1259 |
+
{
|
| 1260 |
+
"epoch": 0.9110512129380054,
|
| 1261 |
+
"grad_norm": 0.5573087930679321,
|
| 1262 |
+
"learning_rate": 0.0005459320021586615,
|
| 1263 |
+
"loss": 3.8402,
|
| 1264 |
+
"step": 8450
|
| 1265 |
+
},
|
| 1266 |
+
{
|
| 1267 |
+
"epoch": 0.9164420485175202,
|
| 1268 |
+
"grad_norm": 0.5278663039207458,
|
| 1269 |
+
"learning_rate": 0.0005456082029141932,
|
| 1270 |
+
"loss": 3.8521,
|
| 1271 |
+
"step": 8500
|
| 1272 |
+
},
|
| 1273 |
+
{
|
| 1274 |
+
"epoch": 0.921832884097035,
|
| 1275 |
+
"grad_norm": 0.6010989546775818,
|
| 1276 |
+
"learning_rate": 0.0005452844036697248,
|
| 1277 |
+
"loss": 3.8536,
|
| 1278 |
+
"step": 8550
|
| 1279 |
+
},
|
| 1280 |
+
{
|
| 1281 |
+
"epoch": 0.9272237196765498,
|
| 1282 |
+
"grad_norm": 0.6041896939277649,
|
| 1283 |
+
"learning_rate": 0.0005449606044252563,
|
| 1284 |
+
"loss": 3.8816,
|
| 1285 |
+
"step": 8600
|
| 1286 |
+
},
|
| 1287 |
+
{
|
| 1288 |
+
"epoch": 0.9326145552560647,
|
| 1289 |
+
"grad_norm": 0.6230564117431641,
|
| 1290 |
+
"learning_rate": 0.0005446368051807879,
|
| 1291 |
+
"loss": 3.8547,
|
| 1292 |
+
"step": 8650
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 0.9380053908355795,
|
| 1296 |
+
"grad_norm": 0.6061603426933289,
|
| 1297 |
+
"learning_rate": 0.0005443130059363194,
|
| 1298 |
+
"loss": 3.8612,
|
| 1299 |
+
"step": 8700
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"epoch": 0.9433962264150944,
|
| 1303 |
+
"grad_norm": 0.532725989818573,
|
| 1304 |
+
"learning_rate": 0.0005439892066918511,
|
| 1305 |
+
"loss": 3.8453,
|
| 1306 |
+
"step": 8750
|
| 1307 |
+
},
|
| 1308 |
+
{
|
| 1309 |
+
"epoch": 0.9487870619946092,
|
| 1310 |
+
"grad_norm": 0.6650099158287048,
|
| 1311 |
+
"learning_rate": 0.0005436654074473825,
|
| 1312 |
+
"loss": 3.845,
|
| 1313 |
+
"step": 8800
|
| 1314 |
+
},
|
| 1315 |
+
{
|
| 1316 |
+
"epoch": 0.954177897574124,
|
| 1317 |
+
"grad_norm": 0.5258676409721375,
|
| 1318 |
+
"learning_rate": 0.0005433416082029142,
|
| 1319 |
+
"loss": 3.8474,
|
| 1320 |
+
"step": 8850
|
| 1321 |
+
},
|
| 1322 |
+
{
|
| 1323 |
+
"epoch": 0.9595687331536388,
|
| 1324 |
+
"grad_norm": 0.5918166637420654,
|
| 1325 |
+
"learning_rate": 0.0005430178089584457,
|
| 1326 |
+
"loss": 3.8476,
|
| 1327 |
+
"step": 8900
|
| 1328 |
+
},
|
| 1329 |
+
{
|
| 1330 |
+
"epoch": 0.9649595687331537,
|
| 1331 |
+
"grad_norm": 0.7267642617225647,
|
| 1332 |
+
"learning_rate": 0.0005426940097139773,
|
| 1333 |
+
"loss": 3.8489,
|
| 1334 |
+
"step": 8950
|
| 1335 |
+
},
|
| 1336 |
+
{
|
| 1337 |
+
"epoch": 0.9703504043126685,
|
| 1338 |
+
"grad_norm": 0.639377772808075,
|
| 1339 |
+
"learning_rate": 0.0005423702104695088,
|
| 1340 |
+
"loss": 3.8648,
|
| 1341 |
+
"step": 9000
|
| 1342 |
+
},
|
| 1343 |
+
{
|
| 1344 |
+
"epoch": 0.9703504043126685,
|
| 1345 |
+
"eval_accuracy": 0.3408996306996996,
|
| 1346 |
+
"eval_loss": 3.779827356338501,
|
| 1347 |
+
"eval_runtime": 181.5426,
|
| 1348 |
+
"eval_samples_per_second": 99.211,
|
| 1349 |
+
"eval_steps_per_second": 6.202,
|
| 1350 |
+
"step": 9000
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 0.9757412398921833,
|
| 1354 |
+
"grad_norm": 0.6780883073806763,
|
| 1355 |
+
"learning_rate": 0.0005420464112250404,
|
| 1356 |
+
"loss": 3.8448,
|
| 1357 |
+
"step": 9050
|
| 1358 |
+
},
|
| 1359 |
+
{
|
| 1360 |
+
"epoch": 0.9811320754716981,
|
| 1361 |
+
"grad_norm": 0.5915326476097107,
|
| 1362 |
+
"learning_rate": 0.000541722611980572,
|
| 1363 |
+
"loss": 3.8335,
|
| 1364 |
+
"step": 9100
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 0.9865229110512129,
|
| 1368 |
+
"grad_norm": 0.5880796313285828,
|
| 1369 |
+
"learning_rate": 0.0005413988127361036,
|
| 1370 |
+
"loss": 3.8434,
|
| 1371 |
+
"step": 9150
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 0.9919137466307277,
|
| 1375 |
+
"grad_norm": 0.551231861114502,
|
| 1376 |
+
"learning_rate": 0.0005410750134916351,
|
| 1377 |
+
"loss": 3.8469,
|
| 1378 |
+
"step": 9200
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 0.9973045822102425,
|
| 1382 |
+
"grad_norm": 0.542789101600647,
|
| 1383 |
+
"learning_rate": 0.0005407512142471667,
|
| 1384 |
+
"loss": 3.8382,
|
| 1385 |
+
"step": 9250
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 1.0026954177897573,
|
| 1389 |
+
"grad_norm": 0.5782141089439392,
|
| 1390 |
+
"learning_rate": 0.0005404274150026983,
|
| 1391 |
+
"loss": 3.8143,
|
| 1392 |
+
"step": 9300
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 1.0080862533692723,
|
| 1396 |
+
"grad_norm": 0.585417628288269,
|
| 1397 |
+
"learning_rate": 0.0005401036157582299,
|
| 1398 |
+
"loss": 3.782,
|
| 1399 |
+
"step": 9350
|
| 1400 |
+
},
|
| 1401 |
+
{
|
| 1402 |
+
"epoch": 1.013477088948787,
|
| 1403 |
+
"grad_norm": 0.5786333680152893,
|
| 1404 |
+
"learning_rate": 0.0005397798165137614,
|
| 1405 |
+
"loss": 3.7919,
|
| 1406 |
+
"step": 9400
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 1.0188679245283019,
|
| 1410 |
+
"grad_norm": 0.5752071738243103,
|
| 1411 |
+
"learning_rate": 0.000539456017269293,
|
| 1412 |
+
"loss": 3.7809,
|
| 1413 |
+
"step": 9450
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 1.0242587601078168,
|
| 1417 |
+
"grad_norm": 0.5453478693962097,
|
| 1418 |
+
"learning_rate": 0.0005391322180248245,
|
| 1419 |
+
"loss": 3.7775,
|
| 1420 |
+
"step": 9500
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 1.0296495956873315,
|
| 1424 |
+
"grad_norm": 0.5964511632919312,
|
| 1425 |
+
"learning_rate": 0.0005388084187803561,
|
| 1426 |
+
"loss": 3.7881,
|
| 1427 |
+
"step": 9550
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 1.0350404312668464,
|
| 1431 |
+
"grad_norm": 0.6188067197799683,
|
| 1432 |
+
"learning_rate": 0.0005384846195358876,
|
| 1433 |
+
"loss": 3.7672,
|
| 1434 |
+
"step": 9600
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 1.0404312668463611,
|
| 1438 |
+
"grad_norm": 0.5592512488365173,
|
| 1439 |
+
"learning_rate": 0.0005381608202914193,
|
| 1440 |
+
"loss": 3.7615,
|
| 1441 |
+
"step": 9650
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 1.045822102425876,
|
| 1445 |
+
"grad_norm": 0.5626137256622314,
|
| 1446 |
+
"learning_rate": 0.0005378370210469509,
|
| 1447 |
+
"loss": 3.7723,
|
| 1448 |
+
"step": 9700
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 1.0512129380053907,
|
| 1452 |
+
"grad_norm": 0.6109785437583923,
|
| 1453 |
+
"learning_rate": 0.0005375132218024824,
|
| 1454 |
+
"loss": 3.7644,
|
| 1455 |
+
"step": 9750
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"epoch": 1.0566037735849056,
|
| 1459 |
+
"grad_norm": 0.6150305271148682,
|
| 1460 |
+
"learning_rate": 0.000537189422558014,
|
| 1461 |
+
"loss": 3.7717,
|
| 1462 |
+
"step": 9800
|
| 1463 |
+
},
|
| 1464 |
+
{
|
| 1465 |
+
"epoch": 1.0619946091644206,
|
| 1466 |
+
"grad_norm": 0.5786086320877075,
|
| 1467 |
+
"learning_rate": 0.0005368656233135455,
|
| 1468 |
+
"loss": 3.7537,
|
| 1469 |
+
"step": 9850
|
| 1470 |
+
},
|
| 1471 |
+
{
|
| 1472 |
+
"epoch": 1.0673854447439353,
|
| 1473 |
+
"grad_norm": 0.5935583710670471,
|
| 1474 |
+
"learning_rate": 0.0005365418240690772,
|
| 1475 |
+
"loss": 3.7646,
|
| 1476 |
+
"step": 9900
|
| 1477 |
+
},
|
| 1478 |
+
{
|
| 1479 |
+
"epoch": 1.0727762803234502,
|
| 1480 |
+
"grad_norm": 0.5563629269599915,
|
| 1481 |
+
"learning_rate": 0.0005362180248246087,
|
| 1482 |
+
"loss": 3.7625,
|
| 1483 |
+
"step": 9950
|
| 1484 |
+
},
|
| 1485 |
+
{
|
| 1486 |
+
"epoch": 1.0781671159029649,
|
| 1487 |
+
"grad_norm": 0.5583340525627136,
|
| 1488 |
+
"learning_rate": 0.0005358942255801403,
|
| 1489 |
+
"loss": 3.7661,
|
| 1490 |
+
"step": 10000
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"epoch": 1.0781671159029649,
|
| 1494 |
+
"eval_accuracy": 0.3447591985806456,
|
| 1495 |
+
"eval_loss": 3.7474043369293213,
|
| 1496 |
+
"eval_runtime": 179.3333,
|
| 1497 |
+
"eval_samples_per_second": 100.433,
|
| 1498 |
+
"eval_steps_per_second": 6.279,
|
| 1499 |
+
"step": 10000
|
| 1500 |
+
},
|
| 1501 |
+
{
|
| 1502 |
+
"epoch": 1.0835579514824798,
|
| 1503 |
+
"grad_norm": 0.5278469324111938,
|
| 1504 |
+
"learning_rate": 0.0005355704263356718,
|
| 1505 |
+
"loss": 3.7572,
|
| 1506 |
+
"step": 10050
|
| 1507 |
+
},
|
| 1508 |
+
{
|
| 1509 |
+
"epoch": 1.0889487870619945,
|
| 1510 |
+
"grad_norm": 0.599793553352356,
|
| 1511 |
+
"learning_rate": 0.0005352466270912035,
|
| 1512 |
+
"loss": 3.7482,
|
| 1513 |
+
"step": 10100
|
| 1514 |
+
},
|
| 1515 |
+
{
|
| 1516 |
+
"epoch": 1.0943396226415094,
|
| 1517 |
+
"grad_norm": 0.594103991985321,
|
| 1518 |
+
"learning_rate": 0.000534922827846735,
|
| 1519 |
+
"loss": 3.7811,
|
| 1520 |
+
"step": 10150
|
| 1521 |
+
},
|
| 1522 |
+
{
|
| 1523 |
+
"epoch": 1.0997304582210243,
|
| 1524 |
+
"grad_norm": 0.539659321308136,
|
| 1525 |
+
"learning_rate": 0.0005345990286022666,
|
| 1526 |
+
"loss": 3.7808,
|
| 1527 |
+
"step": 10200
|
| 1528 |
+
},
|
| 1529 |
+
{
|
| 1530 |
+
"epoch": 1.105121293800539,
|
| 1531 |
+
"grad_norm": 0.5949695110321045,
|
| 1532 |
+
"learning_rate": 0.0005342752293577981,
|
| 1533 |
+
"loss": 3.7483,
|
| 1534 |
+
"step": 10250
|
| 1535 |
+
},
|
| 1536 |
+
{
|
| 1537 |
+
"epoch": 1.110512129380054,
|
| 1538 |
+
"grad_norm": 0.6354559659957886,
|
| 1539 |
+
"learning_rate": 0.0005339514301133297,
|
| 1540 |
+
"loss": 3.7743,
|
| 1541 |
+
"step": 10300
|
| 1542 |
+
},
|
| 1543 |
+
{
|
| 1544 |
+
"epoch": 1.1159029649595686,
|
| 1545 |
+
"grad_norm": 0.5751685500144958,
|
| 1546 |
+
"learning_rate": 0.0005336276308688612,
|
| 1547 |
+
"loss": 3.7727,
|
| 1548 |
+
"step": 10350
|
| 1549 |
+
},
|
| 1550 |
+
{
|
| 1551 |
+
"epoch": 1.1212938005390836,
|
| 1552 |
+
"grad_norm": 0.6152482032775879,
|
| 1553 |
+
"learning_rate": 0.0005333038316243928,
|
| 1554 |
+
"loss": 3.7604,
|
| 1555 |
+
"step": 10400
|
| 1556 |
+
},
|
| 1557 |
+
{
|
| 1558 |
+
"epoch": 1.1266846361185983,
|
| 1559 |
+
"grad_norm": 0.6259979009628296,
|
| 1560 |
+
"learning_rate": 0.0005329800323799244,
|
| 1561 |
+
"loss": 3.7242,
|
| 1562 |
+
"step": 10450
|
| 1563 |
+
},
|
| 1564 |
+
{
|
| 1565 |
+
"epoch": 1.1320754716981132,
|
| 1566 |
+
"grad_norm": 0.5735751390457153,
|
| 1567 |
+
"learning_rate": 0.000532656233135456,
|
| 1568 |
+
"loss": 3.75,
|
| 1569 |
+
"step": 10500
|
| 1570 |
+
},
|
| 1571 |
+
{
|
| 1572 |
+
"epoch": 1.137466307277628,
|
| 1573 |
+
"grad_norm": 0.5335408449172974,
|
| 1574 |
+
"learning_rate": 0.0005323324338909875,
|
| 1575 |
+
"loss": 3.7497,
|
| 1576 |
+
"step": 10550
|
| 1577 |
+
},
|
| 1578 |
+
{
|
| 1579 |
+
"epoch": 1.1428571428571428,
|
| 1580 |
+
"grad_norm": 0.5535331964492798,
|
| 1581 |
+
"learning_rate": 0.0005320086346465191,
|
| 1582 |
+
"loss": 3.757,
|
| 1583 |
+
"step": 10600
|
| 1584 |
+
},
|
| 1585 |
+
{
|
| 1586 |
+
"epoch": 1.1482479784366577,
|
| 1587 |
+
"grad_norm": 0.5518361926078796,
|
| 1588 |
+
"learning_rate": 0.0005316848354020507,
|
| 1589 |
+
"loss": 3.7635,
|
| 1590 |
+
"step": 10650
|
| 1591 |
+
},
|
| 1592 |
+
{
|
| 1593 |
+
"epoch": 1.1536388140161726,
|
| 1594 |
+
"grad_norm": 0.6065233945846558,
|
| 1595 |
+
"learning_rate": 0.0005313610361575823,
|
| 1596 |
+
"loss": 3.7482,
|
| 1597 |
+
"step": 10700
|
| 1598 |
+
},
|
| 1599 |
+
{
|
| 1600 |
+
"epoch": 1.1590296495956873,
|
| 1601 |
+
"grad_norm": 0.5510832667350769,
|
| 1602 |
+
"learning_rate": 0.0005310372369131138,
|
| 1603 |
+
"loss": 3.7464,
|
| 1604 |
+
"step": 10750
|
| 1605 |
+
},
|
| 1606 |
+
{
|
| 1607 |
+
"epoch": 1.1644204851752022,
|
| 1608 |
+
"grad_norm": 0.5745865702629089,
|
| 1609 |
+
"learning_rate": 0.0005307134376686454,
|
| 1610 |
+
"loss": 3.7237,
|
| 1611 |
+
"step": 10800
|
| 1612 |
+
},
|
| 1613 |
+
{
|
| 1614 |
+
"epoch": 1.169811320754717,
|
| 1615 |
+
"grad_norm": 0.6756680011749268,
|
| 1616 |
+
"learning_rate": 0.000530389638424177,
|
| 1617 |
+
"loss": 3.7627,
|
| 1618 |
+
"step": 10850
|
| 1619 |
+
},
|
| 1620 |
+
{
|
| 1621 |
+
"epoch": 1.1752021563342319,
|
| 1622 |
+
"grad_norm": 0.6508898138999939,
|
| 1623 |
+
"learning_rate": 0.0005300658391797085,
|
| 1624 |
+
"loss": 3.7616,
|
| 1625 |
+
"step": 10900
|
| 1626 |
+
},
|
| 1627 |
+
{
|
| 1628 |
+
"epoch": 1.1805929919137466,
|
| 1629 |
+
"grad_norm": 0.6372632384300232,
|
| 1630 |
+
"learning_rate": 0.00052974203993524,
|
| 1631 |
+
"loss": 3.7205,
|
| 1632 |
+
"step": 10950
|
| 1633 |
+
},
|
| 1634 |
+
{
|
| 1635 |
+
"epoch": 1.1859838274932615,
|
| 1636 |
+
"grad_norm": 0.584235668182373,
|
| 1637 |
+
"learning_rate": 0.0005294182406907717,
|
| 1638 |
+
"loss": 3.7415,
|
| 1639 |
+
"step": 11000
|
| 1640 |
+
},
|
| 1641 |
+
{
|
| 1642 |
+
"epoch": 1.1859838274932615,
|
| 1643 |
+
"eval_accuracy": 0.3469530090385078,
|
| 1644 |
+
"eval_loss": 3.7233991622924805,
|
| 1645 |
+
"eval_runtime": 179.2943,
|
| 1646 |
+
"eval_samples_per_second": 100.455,
|
| 1647 |
+
"eval_steps_per_second": 6.28,
|
| 1648 |
+
"step": 11000
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 1.1913746630727764,
|
| 1652 |
+
"grad_norm": 0.5607488751411438,
|
| 1653 |
+
"learning_rate": 0.0005290944414463033,
|
| 1654 |
+
"loss": 3.7504,
|
| 1655 |
+
"step": 11050
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 1.196765498652291,
|
| 1659 |
+
"grad_norm": 0.6091164946556091,
|
| 1660 |
+
"learning_rate": 0.0005287706422018348,
|
| 1661 |
+
"loss": 3.744,
|
| 1662 |
+
"step": 11100
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 1.202156334231806,
|
| 1666 |
+
"grad_norm": 0.6206035614013672,
|
| 1667 |
+
"learning_rate": 0.0005284468429573664,
|
| 1668 |
+
"loss": 3.7378,
|
| 1669 |
+
"step": 11150
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 1.2075471698113207,
|
| 1673 |
+
"grad_norm": 0.6528010964393616,
|
| 1674 |
+
"learning_rate": 0.0005281230437128979,
|
| 1675 |
+
"loss": 3.7509,
|
| 1676 |
+
"step": 11200
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 1.2129380053908356,
|
| 1680 |
+
"grad_norm": 0.520330011844635,
|
| 1681 |
+
"learning_rate": 0.0005278057204533189,
|
| 1682 |
+
"loss": 3.7608,
|
| 1683 |
+
"step": 11250
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 1.2183288409703503,
|
| 1687 |
+
"grad_norm": 0.6227236390113831,
|
| 1688 |
+
"learning_rate": 0.0005274819212088505,
|
| 1689 |
+
"loss": 3.7286,
|
| 1690 |
+
"step": 11300
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 1.2237196765498652,
|
| 1694 |
+
"grad_norm": 0.5682984590530396,
|
| 1695 |
+
"learning_rate": 0.0005271581219643821,
|
| 1696 |
+
"loss": 3.7505,
|
| 1697 |
+
"step": 11350
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 1.2291105121293802,
|
| 1701 |
+
"grad_norm": 0.6123554706573486,
|
| 1702 |
+
"learning_rate": 0.0005268343227199136,
|
| 1703 |
+
"loss": 3.7508,
|
| 1704 |
+
"step": 11400
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 1.2345013477088949,
|
| 1708 |
+
"grad_norm": 0.507030189037323,
|
| 1709 |
+
"learning_rate": 0.0005265105234754452,
|
| 1710 |
+
"loss": 3.7553,
|
| 1711 |
+
"step": 11450
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 1.2398921832884098,
|
| 1715 |
+
"grad_norm": 0.5759048461914062,
|
| 1716 |
+
"learning_rate": 0.0005261867242309767,
|
| 1717 |
+
"loss": 3.7192,
|
| 1718 |
+
"step": 11500
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 1.2452830188679245,
|
| 1722 |
+
"grad_norm": 0.5370573997497559,
|
| 1723 |
+
"learning_rate": 0.0005258629249865083,
|
| 1724 |
+
"loss": 3.7409,
|
| 1725 |
+
"step": 11550
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 1.2506738544474394,
|
| 1729 |
+
"grad_norm": 0.5597508549690247,
|
| 1730 |
+
"learning_rate": 0.0005255391257420398,
|
| 1731 |
+
"loss": 3.739,
|
| 1732 |
+
"step": 11600
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 1.256064690026954,
|
| 1736 |
+
"grad_norm": 0.6163648366928101,
|
| 1737 |
+
"learning_rate": 0.0005252153264975715,
|
| 1738 |
+
"loss": 3.7364,
|
| 1739 |
+
"step": 11650
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 1.261455525606469,
|
| 1743 |
+
"grad_norm": 0.561890721321106,
|
| 1744 |
+
"learning_rate": 0.000524891527253103,
|
| 1745 |
+
"loss": 3.7371,
|
| 1746 |
+
"step": 11700
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 1.266846361185984,
|
| 1750 |
+
"grad_norm": 0.6164953708648682,
|
| 1751 |
+
"learning_rate": 0.0005245677280086346,
|
| 1752 |
+
"loss": 3.7437,
|
| 1753 |
+
"step": 11750
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 1.2722371967654986,
|
| 1757 |
+
"grad_norm": 0.5438469648361206,
|
| 1758 |
+
"learning_rate": 0.0005242439287641661,
|
| 1759 |
+
"loss": 3.7317,
|
| 1760 |
+
"step": 11800
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 1.2776280323450135,
|
| 1764 |
+
"grad_norm": 0.5996424555778503,
|
| 1765 |
+
"learning_rate": 0.0005239201295196978,
|
| 1766 |
+
"loss": 3.7139,
|
| 1767 |
+
"step": 11850
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 1.2830188679245282,
|
| 1771 |
+
"grad_norm": 0.5443153381347656,
|
| 1772 |
+
"learning_rate": 0.0005235963302752293,
|
| 1773 |
+
"loss": 3.7189,
|
| 1774 |
+
"step": 11900
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 1.2884097035040432,
|
| 1778 |
+
"grad_norm": 0.5056083798408508,
|
| 1779 |
+
"learning_rate": 0.0005232725310307609,
|
| 1780 |
+
"loss": 3.7286,
|
| 1781 |
+
"step": 11950
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 1.2938005390835579,
|
| 1785 |
+
"grad_norm": 0.5346047878265381,
|
| 1786 |
+
"learning_rate": 0.0005229487317862924,
|
| 1787 |
+
"loss": 3.7241,
|
| 1788 |
+
"step": 12000
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 1.2938005390835579,
|
| 1792 |
+
"eval_accuracy": 0.3486271327339533,
|
| 1793 |
+
"eval_loss": 3.701392889022827,
|
| 1794 |
+
"eval_runtime": 179.3961,
|
| 1795 |
+
"eval_samples_per_second": 100.398,
|
| 1796 |
+
"eval_steps_per_second": 6.277,
|
| 1797 |
+
"step": 12000
|
| 1798 |
+
},
|
| 1799 |
+
{
|
| 1800 |
+
"epoch": 1.2991913746630728,
|
| 1801 |
+
"grad_norm": 0.561946451663971,
|
| 1802 |
+
"learning_rate": 0.000522624932541824,
|
| 1803 |
+
"loss": 3.7065,
|
| 1804 |
+
"step": 12050
|
| 1805 |
+
},
|
| 1806 |
+
{
|
| 1807 |
+
"epoch": 1.3045822102425877,
|
| 1808 |
+
"grad_norm": 0.6094053387641907,
|
| 1809 |
+
"learning_rate": 0.0005223011332973557,
|
| 1810 |
+
"loss": 3.7432,
|
| 1811 |
+
"step": 12100
|
| 1812 |
+
},
|
| 1813 |
+
{
|
| 1814 |
+
"epoch": 1.3099730458221024,
|
| 1815 |
+
"grad_norm": 0.548875093460083,
|
| 1816 |
+
"learning_rate": 0.0005219773340528872,
|
| 1817 |
+
"loss": 3.7251,
|
| 1818 |
+
"step": 12150
|
| 1819 |
+
},
|
| 1820 |
+
{
|
| 1821 |
+
"epoch": 1.3153638814016173,
|
| 1822 |
+
"grad_norm": 0.5728870630264282,
|
| 1823 |
+
"learning_rate": 0.0005216535348084188,
|
| 1824 |
+
"loss": 3.7107,
|
| 1825 |
+
"step": 12200
|
| 1826 |
+
},
|
| 1827 |
+
{
|
| 1828 |
+
"epoch": 1.320754716981132,
|
| 1829 |
+
"grad_norm": 0.6108008623123169,
|
| 1830 |
+
"learning_rate": 0.0005213297355639503,
|
| 1831 |
+
"loss": 3.7369,
|
| 1832 |
+
"step": 12250
|
| 1833 |
+
},
|
| 1834 |
+
{
|
| 1835 |
+
"epoch": 1.326145552560647,
|
| 1836 |
+
"grad_norm": 0.5879302024841309,
|
| 1837 |
+
"learning_rate": 0.0005210059363194819,
|
| 1838 |
+
"loss": 3.7043,
|
| 1839 |
+
"step": 12300
|
| 1840 |
+
},
|
| 1841 |
+
{
|
| 1842 |
+
"epoch": 1.3315363881401616,
|
| 1843 |
+
"grad_norm": 0.550482988357544,
|
| 1844 |
+
"learning_rate": 0.0005206821370750134,
|
| 1845 |
+
"loss": 3.7288,
|
| 1846 |
+
"step": 12350
|
| 1847 |
+
},
|
| 1848 |
+
{
|
| 1849 |
+
"epoch": 1.3369272237196765,
|
| 1850 |
+
"grad_norm": 0.5876588821411133,
|
| 1851 |
+
"learning_rate": 0.000520358337830545,
|
| 1852 |
+
"loss": 3.7,
|
| 1853 |
+
"step": 12400
|
| 1854 |
+
},
|
| 1855 |
+
{
|
| 1856 |
+
"epoch": 1.3423180592991915,
|
| 1857 |
+
"grad_norm": 0.5548244118690491,
|
| 1858 |
+
"learning_rate": 0.0005200345385860766,
|
| 1859 |
+
"loss": 3.7202,
|
| 1860 |
+
"step": 12450
|
| 1861 |
+
},
|
| 1862 |
+
{
|
| 1863 |
+
"epoch": 1.3477088948787062,
|
| 1864 |
+
"grad_norm": 0.6892051696777344,
|
| 1865 |
+
"learning_rate": 0.0005197107393416082,
|
| 1866 |
+
"loss": 3.7287,
|
| 1867 |
+
"step": 12500
|
| 1868 |
+
},
|
| 1869 |
+
{
|
| 1870 |
+
"epoch": 1.353099730458221,
|
| 1871 |
+
"grad_norm": 0.5671435594558716,
|
| 1872 |
+
"learning_rate": 0.0005193869400971397,
|
| 1873 |
+
"loss": 3.7361,
|
| 1874 |
+
"step": 12550
|
| 1875 |
+
},
|
| 1876 |
+
{
|
| 1877 |
+
"epoch": 1.3584905660377358,
|
| 1878 |
+
"grad_norm": 0.5692296624183655,
|
| 1879 |
+
"learning_rate": 0.0005190631408526713,
|
| 1880 |
+
"loss": 3.7137,
|
| 1881 |
+
"step": 12600
|
| 1882 |
+
},
|
| 1883 |
+
{
|
| 1884 |
+
"epoch": 1.3638814016172507,
|
| 1885 |
+
"grad_norm": 0.5560773611068726,
|
| 1886 |
+
"learning_rate": 0.0005187393416082029,
|
| 1887 |
+
"loss": 3.7092,
|
| 1888 |
+
"step": 12650
|
| 1889 |
+
},
|
| 1890 |
+
{
|
| 1891 |
+
"epoch": 1.3692722371967654,
|
| 1892 |
+
"grad_norm": 0.5214106440544128,
|
| 1893 |
+
"learning_rate": 0.0005184155423637345,
|
| 1894 |
+
"loss": 3.7076,
|
| 1895 |
+
"step": 12700
|
| 1896 |
+
},
|
| 1897 |
+
{
|
| 1898 |
+
"epoch": 1.3746630727762803,
|
| 1899 |
+
"grad_norm": 0.5432198643684387,
|
| 1900 |
+
"learning_rate": 0.000518091743119266,
|
| 1901 |
+
"loss": 3.7224,
|
| 1902 |
+
"step": 12750
|
| 1903 |
+
},
|
| 1904 |
+
{
|
| 1905 |
+
"epoch": 1.3800539083557952,
|
| 1906 |
+
"grad_norm": 0.5707585215568542,
|
| 1907 |
+
"learning_rate": 0.0005177679438747976,
|
| 1908 |
+
"loss": 3.7136,
|
| 1909 |
+
"step": 12800
|
| 1910 |
+
},
|
| 1911 |
+
{
|
| 1912 |
+
"epoch": 1.38544474393531,
|
| 1913 |
+
"grad_norm": 0.583604633808136,
|
| 1914 |
+
"learning_rate": 0.0005174441446303291,
|
| 1915 |
+
"loss": 3.7356,
|
| 1916 |
+
"step": 12850
|
| 1917 |
+
},
|
| 1918 |
+
{
|
| 1919 |
+
"epoch": 1.3908355795148248,
|
| 1920 |
+
"grad_norm": 0.567251443862915,
|
| 1921 |
+
"learning_rate": 0.0005171203453858607,
|
| 1922 |
+
"loss": 3.7341,
|
| 1923 |
+
"step": 12900
|
| 1924 |
+
},
|
| 1925 |
+
{
|
| 1926 |
+
"epoch": 1.3962264150943398,
|
| 1927 |
+
"grad_norm": 0.5615735054016113,
|
| 1928 |
+
"learning_rate": 0.0005167965461413922,
|
| 1929 |
+
"loss": 3.7331,
|
| 1930 |
+
"step": 12950
|
| 1931 |
+
},
|
| 1932 |
+
{
|
| 1933 |
+
"epoch": 1.4016172506738545,
|
| 1934 |
+
"grad_norm": 0.6066398024559021,
|
| 1935 |
+
"learning_rate": 0.0005164727468969239,
|
| 1936 |
+
"loss": 3.715,
|
| 1937 |
+
"step": 13000
|
| 1938 |
+
},
|
| 1939 |
+
{
|
| 1940 |
+
"epoch": 1.4016172506738545,
|
| 1941 |
+
"eval_accuracy": 0.3512098118773035,
|
| 1942 |
+
"eval_loss": 3.67993426322937,
|
| 1943 |
+
"eval_runtime": 181.5275,
|
| 1944 |
+
"eval_samples_per_second": 99.219,
|
| 1945 |
+
"eval_steps_per_second": 6.203,
|
| 1946 |
+
"step": 13000
|
| 1947 |
+
},
|
| 1948 |
+
{
|
| 1949 |
+
"epoch": 1.4070080862533692,
|
| 1950 |
+
"grad_norm": 0.6397793292999268,
|
| 1951 |
+
"learning_rate": 0.0005161489476524554,
|
| 1952 |
+
"loss": 3.6967,
|
| 1953 |
+
"step": 13050
|
| 1954 |
+
},
|
| 1955 |
+
{
|
| 1956 |
+
"epoch": 1.412398921832884,
|
| 1957 |
+
"grad_norm": 0.6502065062522888,
|
| 1958 |
+
"learning_rate": 0.000515825148407987,
|
| 1959 |
+
"loss": 3.7131,
|
| 1960 |
+
"step": 13100
|
| 1961 |
+
},
|
| 1962 |
+
{
|
| 1963 |
+
"epoch": 1.417789757412399,
|
| 1964 |
+
"grad_norm": 0.5458613038063049,
|
| 1965 |
+
"learning_rate": 0.0005155013491635185,
|
| 1966 |
+
"loss": 3.7147,
|
| 1967 |
+
"step": 13150
|
| 1968 |
+
},
|
| 1969 |
+
{
|
| 1970 |
+
"epoch": 1.4231805929919137,
|
| 1971 |
+
"grad_norm": 0.5654726028442383,
|
| 1972 |
+
"learning_rate": 0.0005151775499190501,
|
| 1973 |
+
"loss": 3.7392,
|
| 1974 |
+
"step": 13200
|
| 1975 |
+
},
|
| 1976 |
+
{
|
| 1977 |
+
"epoch": 1.4285714285714286,
|
| 1978 |
+
"grad_norm": 0.5382741689682007,
|
| 1979 |
+
"learning_rate": 0.0005148537506745818,
|
| 1980 |
+
"loss": 3.6966,
|
| 1981 |
+
"step": 13250
|
| 1982 |
+
},
|
| 1983 |
+
{
|
| 1984 |
+
"epoch": 1.4339622641509435,
|
| 1985 |
+
"grad_norm": 0.5260798931121826,
|
| 1986 |
+
"learning_rate": 0.0005145299514301133,
|
| 1987 |
+
"loss": 3.703,
|
| 1988 |
+
"step": 13300
|
| 1989 |
+
},
|
| 1990 |
+
{
|
| 1991 |
+
"epoch": 1.4393530997304582,
|
| 1992 |
+
"grad_norm": 0.5853448510169983,
|
| 1993 |
+
"learning_rate": 0.0005142061521856449,
|
| 1994 |
+
"loss": 3.684,
|
| 1995 |
+
"step": 13350
|
| 1996 |
+
},
|
| 1997 |
+
{
|
| 1998 |
+
"epoch": 1.444743935309973,
|
| 1999 |
+
"grad_norm": 0.538550615310669,
|
| 2000 |
+
"learning_rate": 0.0005138823529411764,
|
| 2001 |
+
"loss": 3.7219,
|
| 2002 |
+
"step": 13400
|
| 2003 |
+
},
|
| 2004 |
+
{
|
| 2005 |
+
"epoch": 1.4501347708894878,
|
| 2006 |
+
"grad_norm": 0.596662700176239,
|
| 2007 |
+
"learning_rate": 0.0005135585536967081,
|
| 2008 |
+
"loss": 3.7194,
|
| 2009 |
+
"step": 13450
|
| 2010 |
+
},
|
| 2011 |
+
{
|
| 2012 |
+
"epoch": 1.4555256064690028,
|
| 2013 |
+
"grad_norm": 0.6186038851737976,
|
| 2014 |
+
"learning_rate": 0.0005132347544522396,
|
| 2015 |
+
"loss": 3.7033,
|
| 2016 |
+
"step": 13500
|
| 2017 |
+
},
|
| 2018 |
+
{
|
| 2019 |
+
"epoch": 1.4609164420485174,
|
| 2020 |
+
"grad_norm": 0.542559802532196,
|
| 2021 |
+
"learning_rate": 0.0005129109552077712,
|
| 2022 |
+
"loss": 3.7194,
|
| 2023 |
+
"step": 13550
|
| 2024 |
+
},
|
| 2025 |
+
{
|
| 2026 |
+
"epoch": 1.4663072776280324,
|
| 2027 |
+
"grad_norm": 0.5352370738983154,
|
| 2028 |
+
"learning_rate": 0.0005125871559633027,
|
| 2029 |
+
"loss": 3.7046,
|
| 2030 |
+
"step": 13600
|
| 2031 |
+
},
|
| 2032 |
+
{
|
| 2033 |
+
"epoch": 1.4716981132075473,
|
| 2034 |
+
"grad_norm": 0.5471335649490356,
|
| 2035 |
+
"learning_rate": 0.0005122633567188343,
|
| 2036 |
+
"loss": 3.7091,
|
| 2037 |
+
"step": 13650
|
| 2038 |
+
},
|
| 2039 |
+
{
|
| 2040 |
+
"epoch": 1.477088948787062,
|
| 2041 |
+
"grad_norm": 0.6399145126342773,
|
| 2042 |
+
"learning_rate": 0.0005119395574743658,
|
| 2043 |
+
"loss": 3.7208,
|
| 2044 |
+
"step": 13700
|
| 2045 |
+
},
|
| 2046 |
+
{
|
| 2047 |
+
"epoch": 1.482479784366577,
|
| 2048 |
+
"grad_norm": 0.5687584280967712,
|
| 2049 |
+
"learning_rate": 0.0005116222342147868,
|
| 2050 |
+
"loss": 3.7,
|
| 2051 |
+
"step": 13750
|
| 2052 |
+
},
|
| 2053 |
+
{
|
| 2054 |
+
"epoch": 1.4878706199460916,
|
| 2055 |
+
"grad_norm": 0.5827769041061401,
|
| 2056 |
+
"learning_rate": 0.0005112984349703183,
|
| 2057 |
+
"loss": 3.7004,
|
| 2058 |
+
"step": 13800
|
| 2059 |
+
},
|
| 2060 |
+
{
|
| 2061 |
+
"epoch": 1.4932614555256065,
|
| 2062 |
+
"grad_norm": 0.5126985907554626,
|
| 2063 |
+
"learning_rate": 0.00051097463572585,
|
| 2064 |
+
"loss": 3.6955,
|
| 2065 |
+
"step": 13850
|
| 2066 |
+
},
|
| 2067 |
+
{
|
| 2068 |
+
"epoch": 1.4986522911051212,
|
| 2069 |
+
"grad_norm": 0.6048992872238159,
|
| 2070 |
+
"learning_rate": 0.0005106508364813815,
|
| 2071 |
+
"loss": 3.6853,
|
| 2072 |
+
"step": 13900
|
| 2073 |
+
},
|
| 2074 |
+
{
|
| 2075 |
+
"epoch": 1.5040431266846361,
|
| 2076 |
+
"grad_norm": 0.5209577679634094,
|
| 2077 |
+
"learning_rate": 0.0005103270372369131,
|
| 2078 |
+
"loss": 3.7061,
|
| 2079 |
+
"step": 13950
|
| 2080 |
+
},
|
| 2081 |
+
{
|
| 2082 |
+
"epoch": 1.509433962264151,
|
| 2083 |
+
"grad_norm": 0.558080792427063,
|
| 2084 |
+
"learning_rate": 0.0005100032379924446,
|
| 2085 |
+
"loss": 3.6982,
|
| 2086 |
+
"step": 14000
|
| 2087 |
+
},
|
| 2088 |
+
{
|
| 2089 |
+
"epoch": 1.509433962264151,
|
| 2090 |
+
"eval_accuracy": 0.35338667248466665,
|
| 2091 |
+
"eval_loss": 3.6556079387664795,
|
| 2092 |
+
"eval_runtime": 181.0746,
|
| 2093 |
+
"eval_samples_per_second": 99.467,
|
| 2094 |
+
"eval_steps_per_second": 6.218,
|
| 2095 |
+
"step": 14000
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 1.5148247978436657,
|
| 2099 |
+
"grad_norm": 0.5223732590675354,
|
| 2100 |
+
"learning_rate": 0.0005096794387479762,
|
| 2101 |
+
"loss": 3.6743,
|
| 2102 |
+
"step": 14050
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 1.5202156334231804,
|
| 2106 |
+
"grad_norm": 0.5266443490982056,
|
| 2107 |
+
"learning_rate": 0.0005093556395035078,
|
| 2108 |
+
"loss": 3.7049,
|
| 2109 |
+
"step": 14100
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 1.5256064690026954,
|
| 2113 |
+
"grad_norm": 0.5410740971565247,
|
| 2114 |
+
"learning_rate": 0.0005090318402590394,
|
| 2115 |
+
"loss": 3.6871,
|
| 2116 |
+
"step": 14150
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 1.5309973045822103,
|
| 2120 |
+
"grad_norm": 0.5198672413825989,
|
| 2121 |
+
"learning_rate": 0.0005087080410145709,
|
| 2122 |
+
"loss": 3.6848,
|
| 2123 |
+
"step": 14200
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 1.536388140161725,
|
| 2127 |
+
"grad_norm": 0.5388085246086121,
|
| 2128 |
+
"learning_rate": 0.0005083842417701025,
|
| 2129 |
+
"loss": 3.6925,
|
| 2130 |
+
"step": 14250
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 1.54177897574124,
|
| 2134 |
+
"grad_norm": 0.5680968761444092,
|
| 2135 |
+
"learning_rate": 0.000508060442525634,
|
| 2136 |
+
"loss": 3.7015,
|
| 2137 |
+
"step": 14300
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 1.5471698113207548,
|
| 2141 |
+
"grad_norm": 0.529816746711731,
|
| 2142 |
+
"learning_rate": 0.0005077366432811656,
|
| 2143 |
+
"loss": 3.7089,
|
| 2144 |
+
"step": 14350
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 1.5525606469002695,
|
| 2148 |
+
"grad_norm": 0.588886022567749,
|
| 2149 |
+
"learning_rate": 0.0005074128440366971,
|
| 2150 |
+
"loss": 3.6898,
|
| 2151 |
+
"step": 14400
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 1.5579514824797842,
|
| 2155 |
+
"grad_norm": 0.53922438621521,
|
| 2156 |
+
"learning_rate": 0.0005070890447922288,
|
| 2157 |
+
"loss": 3.7101,
|
| 2158 |
+
"step": 14450
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 1.5633423180592994,
|
| 2162 |
+
"grad_norm": 0.5846419930458069,
|
| 2163 |
+
"learning_rate": 0.0005067652455477604,
|
| 2164 |
+
"loss": 3.6752,
|
| 2165 |
+
"step": 14500
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 1.568733153638814,
|
| 2169 |
+
"grad_norm": 0.5448037981987,
|
| 2170 |
+
"learning_rate": 0.0005064414463032919,
|
| 2171 |
+
"loss": 3.6868,
|
| 2172 |
+
"step": 14550
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 1.5741239892183287,
|
| 2176 |
+
"grad_norm": 0.5100300908088684,
|
| 2177 |
+
"learning_rate": 0.0005061176470588235,
|
| 2178 |
+
"loss": 3.6773,
|
| 2179 |
+
"step": 14600
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 1.5795148247978437,
|
| 2183 |
+
"grad_norm": 0.5662325620651245,
|
| 2184 |
+
"learning_rate": 0.0005057938478143551,
|
| 2185 |
+
"loss": 3.693,
|
| 2186 |
+
"step": 14650
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 1.5849056603773586,
|
| 2190 |
+
"grad_norm": 0.5295085906982422,
|
| 2191 |
+
"learning_rate": 0.0005054700485698867,
|
| 2192 |
+
"loss": 3.6999,
|
| 2193 |
+
"step": 14700
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 1.5902964959568733,
|
| 2197 |
+
"grad_norm": 0.6152496337890625,
|
| 2198 |
+
"learning_rate": 0.0005051462493254182,
|
| 2199 |
+
"loss": 3.6639,
|
| 2200 |
+
"step": 14750
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 1.595687331536388,
|
| 2204 |
+
"grad_norm": 0.5873838067054749,
|
| 2205 |
+
"learning_rate": 0.0005048224500809498,
|
| 2206 |
+
"loss": 3.6879,
|
| 2207 |
+
"step": 14800
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 1.6010781671159031,
|
| 2211 |
+
"grad_norm": 0.5197713375091553,
|
| 2212 |
+
"learning_rate": 0.0005044986508364813,
|
| 2213 |
+
"loss": 3.6784,
|
| 2214 |
+
"step": 14850
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 1.6064690026954178,
|
| 2218 |
+
"grad_norm": 0.5193688869476318,
|
| 2219 |
+
"learning_rate": 0.0005041748515920129,
|
| 2220 |
+
"loss": 3.6756,
|
| 2221 |
+
"step": 14900
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 1.6118598382749325,
|
| 2225 |
+
"grad_norm": 0.5456000566482544,
|
| 2226 |
+
"learning_rate": 0.0005038510523475444,
|
| 2227 |
+
"loss": 3.6745,
|
| 2228 |
+
"step": 14950
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 1.6172506738544474,
|
| 2232 |
+
"grad_norm": 0.565951943397522,
|
| 2233 |
+
"learning_rate": 0.0005035272531030761,
|
| 2234 |
+
"loss": 3.6829,
|
| 2235 |
+
"step": 15000
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 1.6172506738544474,
|
| 2239 |
+
"eval_accuracy": 0.3553477484568302,
|
| 2240 |
+
"eval_loss": 3.6376187801361084,
|
| 2241 |
+
"eval_runtime": 181.6027,
|
| 2242 |
+
"eval_samples_per_second": 99.178,
|
| 2243 |
+
"eval_steps_per_second": 6.2,
|
| 2244 |
+
"step": 15000
|
| 2245 |
+
},
|
| 2246 |
+
{
|
| 2247 |
+
"epoch": 1.6226415094339623,
|
| 2248 |
+
"grad_norm": 0.5930358171463013,
|
| 2249 |
+
"learning_rate": 0.0005032034538586076,
|
| 2250 |
+
"loss": 3.6899,
|
| 2251 |
+
"step": 15050
|
| 2252 |
+
},
|
| 2253 |
+
{
|
| 2254 |
+
"epoch": 1.628032345013477,
|
| 2255 |
+
"grad_norm": 0.5067706108093262,
|
| 2256 |
+
"learning_rate": 0.0005028796546141392,
|
| 2257 |
+
"loss": 3.6659,
|
| 2258 |
+
"step": 15100
|
| 2259 |
+
},
|
| 2260 |
+
{
|
| 2261 |
+
"epoch": 1.633423180592992,
|
| 2262 |
+
"grad_norm": 0.5201679468154907,
|
| 2263 |
+
"learning_rate": 0.0005025558553696707,
|
| 2264 |
+
"loss": 3.6864,
|
| 2265 |
+
"step": 15150
|
| 2266 |
+
},
|
| 2267 |
+
{
|
| 2268 |
+
"epoch": 1.6388140161725069,
|
| 2269 |
+
"grad_norm": 0.5754110217094421,
|
| 2270 |
+
"learning_rate": 0.0005022320561252023,
|
| 2271 |
+
"loss": 3.6606,
|
| 2272 |
+
"step": 15200
|
| 2273 |
+
},
|
| 2274 |
+
{
|
| 2275 |
+
"epoch": 1.6442048517520216,
|
| 2276 |
+
"grad_norm": 0.592934250831604,
|
| 2277 |
+
"learning_rate": 0.0005019082568807339,
|
| 2278 |
+
"loss": 3.7035,
|
| 2279 |
+
"step": 15250
|
| 2280 |
+
},
|
| 2281 |
+
{
|
| 2282 |
+
"epoch": 1.6495956873315363,
|
| 2283 |
+
"grad_norm": 0.5429046154022217,
|
| 2284 |
+
"learning_rate": 0.0005015844576362655,
|
| 2285 |
+
"loss": 3.687,
|
| 2286 |
+
"step": 15300
|
| 2287 |
+
},
|
| 2288 |
+
{
|
| 2289 |
+
"epoch": 1.6549865229110512,
|
| 2290 |
+
"grad_norm": 0.5521410703659058,
|
| 2291 |
+
"learning_rate": 0.000501260658391797,
|
| 2292 |
+
"loss": 3.6814,
|
| 2293 |
+
"step": 15350
|
| 2294 |
+
},
|
| 2295 |
+
{
|
| 2296 |
+
"epoch": 1.6603773584905661,
|
| 2297 |
+
"grad_norm": 0.5523940324783325,
|
| 2298 |
+
"learning_rate": 0.0005009368591473286,
|
| 2299 |
+
"loss": 3.671,
|
| 2300 |
+
"step": 15400
|
| 2301 |
+
},
|
| 2302 |
+
{
|
| 2303 |
+
"epoch": 1.6657681940700808,
|
| 2304 |
+
"grad_norm": 0.5145202875137329,
|
| 2305 |
+
"learning_rate": 0.0005006130599028602,
|
| 2306 |
+
"loss": 3.6823,
|
| 2307 |
+
"step": 15450
|
| 2308 |
+
},
|
| 2309 |
+
{
|
| 2310 |
+
"epoch": 1.6711590296495957,
|
| 2311 |
+
"grad_norm": 0.541299045085907,
|
| 2312 |
+
"learning_rate": 0.0005002892606583918,
|
| 2313 |
+
"loss": 3.6941,
|
| 2314 |
+
"step": 15500
|
| 2315 |
+
},
|
| 2316 |
+
{
|
| 2317 |
+
"epoch": 1.6765498652291106,
|
| 2318 |
+
"grad_norm": 0.578633189201355,
|
| 2319 |
+
"learning_rate": 0.0004999654614139233,
|
| 2320 |
+
"loss": 3.6671,
|
| 2321 |
+
"step": 15550
|
| 2322 |
+
},
|
| 2323 |
+
{
|
| 2324 |
+
"epoch": 1.6819407008086253,
|
| 2325 |
+
"grad_norm": 0.5630112290382385,
|
| 2326 |
+
"learning_rate": 0.0004996416621694549,
|
| 2327 |
+
"loss": 3.6661,
|
| 2328 |
+
"step": 15600
|
| 2329 |
+
},
|
| 2330 |
+
{
|
| 2331 |
+
"epoch": 1.68733153638814,
|
| 2332 |
+
"grad_norm": 0.546072244644165,
|
| 2333 |
+
"learning_rate": 0.0004993178629249864,
|
| 2334 |
+
"loss": 3.663,
|
| 2335 |
+
"step": 15650
|
| 2336 |
+
},
|
| 2337 |
+
{
|
| 2338 |
+
"epoch": 1.692722371967655,
|
| 2339 |
+
"grad_norm": 0.6002097725868225,
|
| 2340 |
+
"learning_rate": 0.000498994063680518,
|
| 2341 |
+
"loss": 3.6783,
|
| 2342 |
+
"step": 15700
|
| 2343 |
+
},
|
| 2344 |
+
{
|
| 2345 |
+
"epoch": 1.6981132075471699,
|
| 2346 |
+
"grad_norm": 0.6249855756759644,
|
| 2347 |
+
"learning_rate": 0.000498676740420939,
|
| 2348 |
+
"loss": 3.6534,
|
| 2349 |
+
"step": 15750
|
| 2350 |
+
},
|
| 2351 |
+
{
|
| 2352 |
+
"epoch": 1.7035040431266846,
|
| 2353 |
+
"grad_norm": 0.5264080166816711,
|
| 2354 |
+
"learning_rate": 0.0004983529411764705,
|
| 2355 |
+
"loss": 3.6742,
|
| 2356 |
+
"step": 15800
|
| 2357 |
+
},
|
| 2358 |
+
{
|
| 2359 |
+
"epoch": 1.7088948787061995,
|
| 2360 |
+
"grad_norm": 0.5488578081130981,
|
| 2361 |
+
"learning_rate": 0.0004980291419320022,
|
| 2362 |
+
"loss": 3.6645,
|
| 2363 |
+
"step": 15850
|
| 2364 |
+
},
|
| 2365 |
+
{
|
| 2366 |
+
"epoch": 1.7142857142857144,
|
| 2367 |
+
"grad_norm": 0.5747875571250916,
|
| 2368 |
+
"learning_rate": 0.0004977053426875337,
|
| 2369 |
+
"loss": 3.6661,
|
| 2370 |
+
"step": 15900
|
| 2371 |
+
},
|
| 2372 |
+
{
|
| 2373 |
+
"epoch": 1.719676549865229,
|
| 2374 |
+
"grad_norm": 0.5004916191101074,
|
| 2375 |
+
"learning_rate": 0.0004973815434430653,
|
| 2376 |
+
"loss": 3.6556,
|
| 2377 |
+
"step": 15950
|
| 2378 |
+
},
|
| 2379 |
+
{
|
| 2380 |
+
"epoch": 1.7250673854447438,
|
| 2381 |
+
"grad_norm": 0.5628464221954346,
|
| 2382 |
+
"learning_rate": 0.0004970577441985968,
|
| 2383 |
+
"loss": 3.6554,
|
| 2384 |
+
"step": 16000
|
| 2385 |
+
},
|
| 2386 |
+
{
|
| 2387 |
+
"epoch": 1.7250673854447438,
|
| 2388 |
+
"eval_accuracy": 0.3569113721653684,
|
| 2389 |
+
"eval_loss": 3.6237761974334717,
|
| 2390 |
+
"eval_runtime": 181.1991,
|
| 2391 |
+
"eval_samples_per_second": 99.399,
|
| 2392 |
+
"eval_steps_per_second": 6.214,
|
| 2393 |
+
"step": 16000
|
| 2394 |
+
},
|
| 2395 |
+
{
|
| 2396 |
+
"epoch": 1.7304582210242587,
|
| 2397 |
+
"grad_norm": 0.5369266271591187,
|
| 2398 |
+
"learning_rate": 0.0004967339449541284,
|
| 2399 |
+
"loss": 3.6758,
|
| 2400 |
+
"step": 16050
|
| 2401 |
+
},
|
| 2402 |
+
{
|
| 2403 |
+
"epoch": 1.7358490566037736,
|
| 2404 |
+
"grad_norm": 0.575958788394928,
|
| 2405 |
+
"learning_rate": 0.00049641014570966,
|
| 2406 |
+
"loss": 3.6766,
|
| 2407 |
+
"step": 16100
|
| 2408 |
+
},
|
| 2409 |
+
{
|
| 2410 |
+
"epoch": 1.7412398921832883,
|
| 2411 |
+
"grad_norm": 0.6171749234199524,
|
| 2412 |
+
"learning_rate": 0.0004960863464651916,
|
| 2413 |
+
"loss": 3.658,
|
| 2414 |
+
"step": 16150
|
| 2415 |
+
},
|
| 2416 |
+
{
|
| 2417 |
+
"epoch": 1.7466307277628033,
|
| 2418 |
+
"grad_norm": 0.5248631238937378,
|
| 2419 |
+
"learning_rate": 0.0004957625472207231,
|
| 2420 |
+
"loss": 3.6652,
|
| 2421 |
+
"step": 16200
|
| 2422 |
+
},
|
| 2423 |
+
{
|
| 2424 |
+
"epoch": 1.7520215633423182,
|
| 2425 |
+
"grad_norm": 0.5663668513298035,
|
| 2426 |
+
"learning_rate": 0.0004954387479762547,
|
| 2427 |
+
"loss": 3.6645,
|
| 2428 |
+
"step": 16250
|
| 2429 |
+
},
|
| 2430 |
+
{
|
| 2431 |
+
"epoch": 1.7574123989218329,
|
| 2432 |
+
"grad_norm": 0.5405056476593018,
|
| 2433 |
+
"learning_rate": 0.0004951149487317862,
|
| 2434 |
+
"loss": 3.6633,
|
| 2435 |
+
"step": 16300
|
| 2436 |
+
},
|
| 2437 |
+
{
|
| 2438 |
+
"epoch": 1.7628032345013476,
|
| 2439 |
+
"grad_norm": 0.5469680428504944,
|
| 2440 |
+
"learning_rate": 0.0004947911494873178,
|
| 2441 |
+
"loss": 3.6642,
|
| 2442 |
+
"step": 16350
|
| 2443 |
+
},
|
| 2444 |
+
{
|
| 2445 |
+
"epoch": 1.7681940700808625,
|
| 2446 |
+
"grad_norm": 0.5379207134246826,
|
| 2447 |
+
"learning_rate": 0.0004944673502428493,
|
| 2448 |
+
"loss": 3.6666,
|
| 2449 |
+
"step": 16400
|
| 2450 |
+
},
|
| 2451 |
+
{
|
| 2452 |
+
"epoch": 1.7735849056603774,
|
| 2453 |
+
"grad_norm": 0.5439175963401794,
|
| 2454 |
+
"learning_rate": 0.000494143550998381,
|
| 2455 |
+
"loss": 3.6703,
|
| 2456 |
+
"step": 16450
|
| 2457 |
+
},
|
| 2458 |
+
{
|
| 2459 |
+
"epoch": 1.778975741239892,
|
| 2460 |
+
"grad_norm": 0.5585680603981018,
|
| 2461 |
+
"learning_rate": 0.0004938197517539125,
|
| 2462 |
+
"loss": 3.666,
|
| 2463 |
+
"step": 16500
|
| 2464 |
+
},
|
| 2465 |
+
{
|
| 2466 |
+
"epoch": 1.784366576819407,
|
| 2467 |
+
"grad_norm": 0.5458288788795471,
|
| 2468 |
+
"learning_rate": 0.0004934959525094441,
|
| 2469 |
+
"loss": 3.6453,
|
| 2470 |
+
"step": 16550
|
| 2471 |
+
},
|
| 2472 |
+
{
|
| 2473 |
+
"epoch": 1.789757412398922,
|
| 2474 |
+
"grad_norm": 0.5485697388648987,
|
| 2475 |
+
"learning_rate": 0.0004931721532649756,
|
| 2476 |
+
"loss": 3.6609,
|
| 2477 |
+
"step": 16600
|
| 2478 |
+
},
|
| 2479 |
+
{
|
| 2480 |
+
"epoch": 1.7951482479784366,
|
| 2481 |
+
"grad_norm": 0.5749149918556213,
|
| 2482 |
+
"learning_rate": 0.0004928483540205073,
|
| 2483 |
+
"loss": 3.6645,
|
| 2484 |
+
"step": 16650
|
| 2485 |
+
},
|
| 2486 |
+
{
|
| 2487 |
+
"epoch": 1.8005390835579513,
|
| 2488 |
+
"grad_norm": 0.5598737001419067,
|
| 2489 |
+
"learning_rate": 0.0004925245547760388,
|
| 2490 |
+
"loss": 3.6582,
|
| 2491 |
+
"step": 16700
|
| 2492 |
+
},
|
| 2493 |
+
{
|
| 2494 |
+
"epoch": 1.8059299191374663,
|
| 2495 |
+
"grad_norm": 0.5720735788345337,
|
| 2496 |
+
"learning_rate": 0.0004922007555315704,
|
| 2497 |
+
"loss": 3.6479,
|
| 2498 |
+
"step": 16750
|
| 2499 |
+
},
|
| 2500 |
+
{
|
| 2501 |
+
"epoch": 1.8113207547169812,
|
| 2502 |
+
"grad_norm": 0.544169008731842,
|
| 2503 |
+
"learning_rate": 0.000491876956287102,
|
| 2504 |
+
"loss": 3.6539,
|
| 2505 |
+
"step": 16800
|
| 2506 |
+
},
|
| 2507 |
+
{
|
| 2508 |
+
"epoch": 1.8167115902964959,
|
| 2509 |
+
"grad_norm": 0.587104082107544,
|
| 2510 |
+
"learning_rate": 0.0004915531570426335,
|
| 2511 |
+
"loss": 3.6497,
|
| 2512 |
+
"step": 16850
|
| 2513 |
+
},
|
| 2514 |
+
{
|
| 2515 |
+
"epoch": 1.8221024258760108,
|
| 2516 |
+
"grad_norm": 0.556251585483551,
|
| 2517 |
+
"learning_rate": 0.0004912293577981652,
|
| 2518 |
+
"loss": 3.6587,
|
| 2519 |
+
"step": 16900
|
| 2520 |
+
},
|
| 2521 |
+
{
|
| 2522 |
+
"epoch": 1.8274932614555257,
|
| 2523 |
+
"grad_norm": 0.5477184057235718,
|
| 2524 |
+
"learning_rate": 0.0004909055585536966,
|
| 2525 |
+
"loss": 3.6421,
|
| 2526 |
+
"step": 16950
|
| 2527 |
+
},
|
| 2528 |
+
{
|
| 2529 |
+
"epoch": 1.8328840970350404,
|
| 2530 |
+
"grad_norm": 0.5440017580986023,
|
| 2531 |
+
"learning_rate": 0.0004905817593092283,
|
| 2532 |
+
"loss": 3.6345,
|
| 2533 |
+
"step": 17000
|
| 2534 |
+
},
|
| 2535 |
+
{
|
| 2536 |
+
"epoch": 1.8328840970350404,
|
| 2537 |
+
"eval_accuracy": 0.3582564949165117,
|
| 2538 |
+
"eval_loss": 3.606245994567871,
|
| 2539 |
+
"eval_runtime": 181.6324,
|
| 2540 |
+
"eval_samples_per_second": 99.162,
|
| 2541 |
+
"eval_steps_per_second": 6.199,
|
| 2542 |
+
"step": 17000
|
| 2543 |
+
},
|
| 2544 |
+
{
|
| 2545 |
+
"epoch": 1.838274932614555,
|
| 2546 |
+
"grad_norm": 0.5022566318511963,
|
| 2547 |
+
"learning_rate": 0.0004902579600647598,
|
| 2548 |
+
"loss": 3.6459,
|
| 2549 |
+
"step": 17050
|
| 2550 |
+
},
|
| 2551 |
+
{
|
| 2552 |
+
"epoch": 1.8436657681940702,
|
| 2553 |
+
"grad_norm": 0.572801947593689,
|
| 2554 |
+
"learning_rate": 0.0004899341608202914,
|
| 2555 |
+
"loss": 3.646,
|
| 2556 |
+
"step": 17100
|
| 2557 |
+
},
|
| 2558 |
+
{
|
| 2559 |
+
"epoch": 1.849056603773585,
|
| 2560 |
+
"grad_norm": 0.5857950448989868,
|
| 2561 |
+
"learning_rate": 0.0004896103615758229,
|
| 2562 |
+
"loss": 3.6343,
|
| 2563 |
+
"step": 17150
|
| 2564 |
+
},
|
| 2565 |
+
{
|
| 2566 |
+
"epoch": 1.8544474393530996,
|
| 2567 |
+
"grad_norm": 0.529326856136322,
|
| 2568 |
+
"learning_rate": 0.0004892865623313546,
|
| 2569 |
+
"loss": 3.6621,
|
| 2570 |
+
"step": 17200
|
| 2571 |
+
},
|
| 2572 |
+
{
|
| 2573 |
+
"epoch": 1.8598382749326146,
|
| 2574 |
+
"grad_norm": 0.5415377616882324,
|
| 2575 |
+
"learning_rate": 0.0004889627630868861,
|
| 2576 |
+
"loss": 3.6692,
|
| 2577 |
+
"step": 17250
|
| 2578 |
+
},
|
| 2579 |
+
{
|
| 2580 |
+
"epoch": 1.8652291105121295,
|
| 2581 |
+
"grad_norm": 0.5492255091667175,
|
| 2582 |
+
"learning_rate": 0.0004886389638424177,
|
| 2583 |
+
"loss": 3.6534,
|
| 2584 |
+
"step": 17300
|
| 2585 |
+
},
|
| 2586 |
+
{
|
| 2587 |
+
"epoch": 1.8706199460916442,
|
| 2588 |
+
"grad_norm": 0.5427126884460449,
|
| 2589 |
+
"learning_rate": 0.0004883151645979492,
|
| 2590 |
+
"loss": 3.6425,
|
| 2591 |
+
"step": 17350
|
| 2592 |
+
},
|
| 2593 |
+
{
|
| 2594 |
+
"epoch": 1.8760107816711589,
|
| 2595 |
+
"grad_norm": 0.561755359172821,
|
| 2596 |
+
"learning_rate": 0.00048799136535348077,
|
| 2597 |
+
"loss": 3.6583,
|
| 2598 |
+
"step": 17400
|
| 2599 |
+
},
|
| 2600 |
+
{
|
| 2601 |
+
"epoch": 1.881401617250674,
|
| 2602 |
+
"grad_norm": 0.5410698056221008,
|
| 2603 |
+
"learning_rate": 0.0004876675661090124,
|
| 2604 |
+
"loss": 3.6306,
|
| 2605 |
+
"step": 17450
|
| 2606 |
+
},
|
| 2607 |
+
{
|
| 2608 |
+
"epoch": 1.8867924528301887,
|
| 2609 |
+
"grad_norm": 0.542617678642273,
|
| 2610 |
+
"learning_rate": 0.0004873437668645439,
|
| 2611 |
+
"loss": 3.6534,
|
| 2612 |
+
"step": 17500
|
| 2613 |
+
},
|
| 2614 |
+
{
|
| 2615 |
+
"epoch": 1.8921832884097034,
|
| 2616 |
+
"grad_norm": 0.5104948878288269,
|
| 2617 |
+
"learning_rate": 0.00048701996762007553,
|
| 2618 |
+
"loss": 3.6543,
|
| 2619 |
+
"step": 17550
|
| 2620 |
+
},
|
| 2621 |
+
{
|
| 2622 |
+
"epoch": 1.8975741239892183,
|
| 2623 |
+
"grad_norm": 0.6261042356491089,
|
| 2624 |
+
"learning_rate": 0.0004866961683756071,
|
| 2625 |
+
"loss": 3.6425,
|
| 2626 |
+
"step": 17600
|
| 2627 |
+
},
|
| 2628 |
+
{
|
| 2629 |
+
"epoch": 1.9029649595687332,
|
| 2630 |
+
"grad_norm": 0.5958898067474365,
|
| 2631 |
+
"learning_rate": 0.0004863723691311387,
|
| 2632 |
+
"loss": 3.6418,
|
| 2633 |
+
"step": 17650
|
| 2634 |
+
},
|
| 2635 |
+
{
|
| 2636 |
+
"epoch": 1.908355795148248,
|
| 2637 |
+
"grad_norm": 0.5822218656539917,
|
| 2638 |
+
"learning_rate": 0.00048604856988667024,
|
| 2639 |
+
"loss": 3.6586,
|
| 2640 |
+
"step": 17700
|
| 2641 |
+
},
|
| 2642 |
+
{
|
| 2643 |
+
"epoch": 1.9137466307277629,
|
| 2644 |
+
"grad_norm": 0.5634023547172546,
|
| 2645 |
+
"learning_rate": 0.0004857247706422018,
|
| 2646 |
+
"loss": 3.6335,
|
| 2647 |
+
"step": 17750
|
| 2648 |
+
},
|
| 2649 |
+
{
|
| 2650 |
+
"epoch": 1.9191374663072778,
|
| 2651 |
+
"grad_norm": 0.5490560531616211,
|
| 2652 |
+
"learning_rate": 0.00048540744738262274,
|
| 2653 |
+
"loss": 3.6368,
|
| 2654 |
+
"step": 17800
|
| 2655 |
+
},
|
| 2656 |
+
{
|
| 2657 |
+
"epoch": 1.9245283018867925,
|
| 2658 |
+
"grad_norm": 0.5121088027954102,
|
| 2659 |
+
"learning_rate": 0.00048508364813815434,
|
| 2660 |
+
"loss": 3.6266,
|
| 2661 |
+
"step": 17850
|
| 2662 |
+
},
|
| 2663 |
+
{
|
| 2664 |
+
"epoch": 1.9299191374663072,
|
| 2665 |
+
"grad_norm": 0.5225231647491455,
|
| 2666 |
+
"learning_rate": 0.00048475984889368584,
|
| 2667 |
+
"loss": 3.6324,
|
| 2668 |
+
"step": 17900
|
| 2669 |
+
},
|
| 2670 |
+
{
|
| 2671 |
+
"epoch": 1.935309973045822,
|
| 2672 |
+
"grad_norm": 0.5291708111763,
|
| 2673 |
+
"learning_rate": 0.0004844360496492175,
|
| 2674 |
+
"loss": 3.6424,
|
| 2675 |
+
"step": 17950
|
| 2676 |
+
},
|
| 2677 |
+
{
|
| 2678 |
+
"epoch": 1.940700808625337,
|
| 2679 |
+
"grad_norm": 0.5437564849853516,
|
| 2680 |
+
"learning_rate": 0.000484112250404749,
|
| 2681 |
+
"loss": 3.6469,
|
| 2682 |
+
"step": 18000
|
| 2683 |
+
},
|
| 2684 |
+
{
|
| 2685 |
+
"epoch": 1.940700808625337,
|
| 2686 |
+
"eval_accuracy": 0.3598998698447057,
|
| 2687 |
+
"eval_loss": 3.5913586616516113,
|
| 2688 |
+
"eval_runtime": 181.4123,
|
| 2689 |
+
"eval_samples_per_second": 99.282,
|
| 2690 |
+
"eval_steps_per_second": 6.207,
|
| 2691 |
+
"step": 18000
|
| 2692 |
+
},
|
| 2693 |
+
{
|
| 2694 |
+
"epoch": 1.9460916442048517,
|
| 2695 |
+
"grad_norm": 0.5686632394790649,
|
| 2696 |
+
"learning_rate": 0.00048378845116028055,
|
| 2697 |
+
"loss": 3.6235,
|
| 2698 |
+
"step": 18050
|
| 2699 |
+
},
|
| 2700 |
+
{
|
| 2701 |
+
"epoch": 1.9514824797843666,
|
| 2702 |
+
"grad_norm": 0.5239707231521606,
|
| 2703 |
+
"learning_rate": 0.00048346465191581215,
|
| 2704 |
+
"loss": 3.6487,
|
| 2705 |
+
"step": 18100
|
| 2706 |
+
},
|
| 2707 |
+
{
|
| 2708 |
+
"epoch": 1.9568733153638815,
|
| 2709 |
+
"grad_norm": 0.5632966756820679,
|
| 2710 |
+
"learning_rate": 0.0004831408526713437,
|
| 2711 |
+
"loss": 3.642,
|
| 2712 |
+
"step": 18150
|
| 2713 |
+
},
|
| 2714 |
+
{
|
| 2715 |
+
"epoch": 1.9622641509433962,
|
| 2716 |
+
"grad_norm": 0.5901452898979187,
|
| 2717 |
+
"learning_rate": 0.0004828170534268753,
|
| 2718 |
+
"loss": 3.6363,
|
| 2719 |
+
"step": 18200
|
| 2720 |
+
},
|
| 2721 |
+
{
|
| 2722 |
+
"epoch": 1.967654986522911,
|
| 2723 |
+
"grad_norm": 0.5481223464012146,
|
| 2724 |
+
"learning_rate": 0.00048249325418240686,
|
| 2725 |
+
"loss": 3.6498,
|
| 2726 |
+
"step": 18250
|
| 2727 |
+
},
|
| 2728 |
+
{
|
| 2729 |
+
"epoch": 1.9730458221024259,
|
| 2730 |
+
"grad_norm": 0.5510320067405701,
|
| 2731 |
+
"learning_rate": 0.00048216945493793846,
|
| 2732 |
+
"loss": 3.6471,
|
| 2733 |
+
"step": 18300
|
| 2734 |
+
},
|
| 2735 |
+
{
|
| 2736 |
+
"epoch": 1.9784366576819408,
|
| 2737 |
+
"grad_norm": 0.5957186222076416,
|
| 2738 |
+
"learning_rate": 0.00048184565569347,
|
| 2739 |
+
"loss": 3.6295,
|
| 2740 |
+
"step": 18350
|
| 2741 |
+
},
|
| 2742 |
+
{
|
| 2743 |
+
"epoch": 1.9838274932614555,
|
| 2744 |
+
"grad_norm": 0.5391411781311035,
|
| 2745 |
+
"learning_rate": 0.0004815218564490016,
|
| 2746 |
+
"loss": 3.6625,
|
| 2747 |
+
"step": 18400
|
| 2748 |
+
},
|
| 2749 |
+
{
|
| 2750 |
+
"epoch": 1.9892183288409704,
|
| 2751 |
+
"grad_norm": 0.5556368231773376,
|
| 2752 |
+
"learning_rate": 0.00048119805720453317,
|
| 2753 |
+
"loss": 3.6166,
|
| 2754 |
+
"step": 18450
|
| 2755 |
+
},
|
| 2756 |
+
{
|
| 2757 |
+
"epoch": 1.9946091644204853,
|
| 2758 |
+
"grad_norm": 0.6190194487571716,
|
| 2759 |
+
"learning_rate": 0.0004808742579600647,
|
| 2760 |
+
"loss": 3.6231,
|
| 2761 |
+
"step": 18500
|
| 2762 |
+
},
|
| 2763 |
+
{
|
| 2764 |
+
"epoch": 2.0,
|
| 2765 |
+
"grad_norm": 1.197678804397583,
|
| 2766 |
+
"learning_rate": 0.0004805504587155963,
|
| 2767 |
+
"loss": 3.6412,
|
| 2768 |
+
"step": 18550
|
| 2769 |
+
},
|
| 2770 |
+
{
|
| 2771 |
+
"epoch": 2.0053908355795147,
|
| 2772 |
+
"grad_norm": 0.5590324401855469,
|
| 2773 |
+
"learning_rate": 0.0004802266594711278,
|
| 2774 |
+
"loss": 3.5499,
|
| 2775 |
+
"step": 18600
|
| 2776 |
+
},
|
| 2777 |
+
{
|
| 2778 |
+
"epoch": 2.01078167115903,
|
| 2779 |
+
"grad_norm": 0.5855118632316589,
|
| 2780 |
+
"learning_rate": 0.0004799028602266594,
|
| 2781 |
+
"loss": 3.5515,
|
| 2782 |
+
"step": 18650
|
| 2783 |
+
},
|
| 2784 |
+
{
|
| 2785 |
+
"epoch": 2.0161725067385445,
|
| 2786 |
+
"grad_norm": 0.5683425664901733,
|
| 2787 |
+
"learning_rate": 0.000479579060982191,
|
| 2788 |
+
"loss": 3.5434,
|
| 2789 |
+
"step": 18700
|
| 2790 |
+
},
|
| 2791 |
+
{
|
| 2792 |
+
"epoch": 2.0215633423180592,
|
| 2793 |
+
"grad_norm": 0.5408341884613037,
|
| 2794 |
+
"learning_rate": 0.0004792552617377226,
|
| 2795 |
+
"loss": 3.5613,
|
| 2796 |
+
"step": 18750
|
| 2797 |
+
},
|
| 2798 |
+
{
|
| 2799 |
+
"epoch": 2.026954177897574,
|
| 2800 |
+
"grad_norm": 0.5905522108078003,
|
| 2801 |
+
"learning_rate": 0.00047893146249325413,
|
| 2802 |
+
"loss": 3.5492,
|
| 2803 |
+
"step": 18800
|
| 2804 |
+
},
|
| 2805 |
+
{
|
| 2806 |
+
"epoch": 2.032345013477089,
|
| 2807 |
+
"grad_norm": 0.5483571887016296,
|
| 2808 |
+
"learning_rate": 0.0004786076632487857,
|
| 2809 |
+
"loss": 3.5309,
|
| 2810 |
+
"step": 18850
|
| 2811 |
+
},
|
| 2812 |
+
{
|
| 2813 |
+
"epoch": 2.0377358490566038,
|
| 2814 |
+
"grad_norm": 0.5604623556137085,
|
| 2815 |
+
"learning_rate": 0.0004782838640043173,
|
| 2816 |
+
"loss": 3.5583,
|
| 2817 |
+
"step": 18900
|
| 2818 |
+
},
|
| 2819 |
+
{
|
| 2820 |
+
"epoch": 2.0431266846361185,
|
| 2821 |
+
"grad_norm": 0.5344943404197693,
|
| 2822 |
+
"learning_rate": 0.00047796006475984883,
|
| 2823 |
+
"loss": 3.5482,
|
| 2824 |
+
"step": 18950
|
| 2825 |
+
},
|
| 2826 |
+
{
|
| 2827 |
+
"epoch": 2.0485175202156336,
|
| 2828 |
+
"grad_norm": 0.5930701494216919,
|
| 2829 |
+
"learning_rate": 0.00047763626551538044,
|
| 2830 |
+
"loss": 3.5639,
|
| 2831 |
+
"step": 19000
|
| 2832 |
+
},
|
| 2833 |
+
{
|
| 2834 |
+
"epoch": 2.0485175202156336,
|
| 2835 |
+
"eval_accuracy": 0.36086101329031256,
|
| 2836 |
+
"eval_loss": 3.584812641143799,
|
| 2837 |
+
"eval_runtime": 181.1798,
|
| 2838 |
+
"eval_samples_per_second": 99.41,
|
| 2839 |
+
"eval_steps_per_second": 6.215,
|
| 2840 |
+
"step": 19000
|
| 2841 |
+
},
|
| 2842 |
+
{
|
| 2843 |
+
"epoch": 2.0539083557951483,
|
| 2844 |
+
"grad_norm": 0.597104012966156,
|
| 2845 |
+
"learning_rate": 0.000477312466270912,
|
| 2846 |
+
"loss": 3.5241,
|
| 2847 |
+
"step": 19050
|
| 2848 |
+
},
|
| 2849 |
+
{
|
| 2850 |
+
"epoch": 2.059299191374663,
|
| 2851 |
+
"grad_norm": 0.5604715943336487,
|
| 2852 |
+
"learning_rate": 0.0004769886670264436,
|
| 2853 |
+
"loss": 3.5646,
|
| 2854 |
+
"step": 19100
|
| 2855 |
+
},
|
| 2856 |
+
{
|
| 2857 |
+
"epoch": 2.0646900269541777,
|
| 2858 |
+
"grad_norm": 0.5276287198066711,
|
| 2859 |
+
"learning_rate": 0.00047666486778197515,
|
| 2860 |
+
"loss": 3.549,
|
| 2861 |
+
"step": 19150
|
| 2862 |
+
},
|
| 2863 |
+
{
|
| 2864 |
+
"epoch": 2.070080862533693,
|
| 2865 |
+
"grad_norm": 0.5561823844909668,
|
| 2866 |
+
"learning_rate": 0.00047634106853750675,
|
| 2867 |
+
"loss": 3.5711,
|
| 2868 |
+
"step": 19200
|
| 2869 |
+
},
|
| 2870 |
+
{
|
| 2871 |
+
"epoch": 2.0754716981132075,
|
| 2872 |
+
"grad_norm": 0.6093335151672363,
|
| 2873 |
+
"learning_rate": 0.00047601726929303825,
|
| 2874 |
+
"loss": 3.5684,
|
| 2875 |
+
"step": 19250
|
| 2876 |
+
},
|
| 2877 |
+
{
|
| 2878 |
+
"epoch": 2.0808625336927222,
|
| 2879 |
+
"grad_norm": 0.588876485824585,
|
| 2880 |
+
"learning_rate": 0.0004756934700485698,
|
| 2881 |
+
"loss": 3.5646,
|
| 2882 |
+
"step": 19300
|
| 2883 |
+
},
|
| 2884 |
+
{
|
| 2885 |
+
"epoch": 2.0862533692722374,
|
| 2886 |
+
"grad_norm": 0.5568276643753052,
|
| 2887 |
+
"learning_rate": 0.0004753696708041014,
|
| 2888 |
+
"loss": 3.5551,
|
| 2889 |
+
"step": 19350
|
| 2890 |
+
},
|
| 2891 |
+
{
|
| 2892 |
+
"epoch": 2.091644204851752,
|
| 2893 |
+
"grad_norm": 0.5360884666442871,
|
| 2894 |
+
"learning_rate": 0.00047504587155963295,
|
| 2895 |
+
"loss": 3.5605,
|
| 2896 |
+
"step": 19400
|
| 2897 |
+
},
|
| 2898 |
+
{
|
| 2899 |
+
"epoch": 2.0970350404312668,
|
| 2900 |
+
"grad_norm": 0.5348809361457825,
|
| 2901 |
+
"learning_rate": 0.00047472207231516456,
|
| 2902 |
+
"loss": 3.5486,
|
| 2903 |
+
"step": 19450
|
| 2904 |
+
},
|
| 2905 |
+
{
|
| 2906 |
+
"epoch": 2.1024258760107815,
|
| 2907 |
+
"grad_norm": 0.6352576017379761,
|
| 2908 |
+
"learning_rate": 0.0004743982730706961,
|
| 2909 |
+
"loss": 3.5433,
|
| 2910 |
+
"step": 19500
|
| 2911 |
+
},
|
| 2912 |
+
{
|
| 2913 |
+
"epoch": 2.1078167115902966,
|
| 2914 |
+
"grad_norm": 0.570564329624176,
|
| 2915 |
+
"learning_rate": 0.0004740744738262277,
|
| 2916 |
+
"loss": 3.5608,
|
| 2917 |
+
"step": 19550
|
| 2918 |
+
},
|
| 2919 |
+
{
|
| 2920 |
+
"epoch": 2.1132075471698113,
|
| 2921 |
+
"grad_norm": 0.5465729832649231,
|
| 2922 |
+
"learning_rate": 0.00047375067458175926,
|
| 2923 |
+
"loss": 3.5578,
|
| 2924 |
+
"step": 19600
|
| 2925 |
+
},
|
| 2926 |
+
{
|
| 2927 |
+
"epoch": 2.118598382749326,
|
| 2928 |
+
"grad_norm": 0.5874576568603516,
|
| 2929 |
+
"learning_rate": 0.00047342687533729087,
|
| 2930 |
+
"loss": 3.5579,
|
| 2931 |
+
"step": 19650
|
| 2932 |
+
},
|
| 2933 |
+
{
|
| 2934 |
+
"epoch": 2.123989218328841,
|
| 2935 |
+
"grad_norm": 0.59136563539505,
|
| 2936 |
+
"learning_rate": 0.0004731030760928224,
|
| 2937 |
+
"loss": 3.5486,
|
| 2938 |
+
"step": 19700
|
| 2939 |
+
},
|
| 2940 |
+
{
|
| 2941 |
+
"epoch": 2.129380053908356,
|
| 2942 |
+
"grad_norm": 0.6302415728569031,
|
| 2943 |
+
"learning_rate": 0.00047277927684835397,
|
| 2944 |
+
"loss": 3.5422,
|
| 2945 |
+
"step": 19750
|
| 2946 |
+
},
|
| 2947 |
+
{
|
| 2948 |
+
"epoch": 2.1347708894878705,
|
| 2949 |
+
"grad_norm": 0.544438898563385,
|
| 2950 |
+
"learning_rate": 0.0004724619535887749,
|
| 2951 |
+
"loss": 3.5742,
|
| 2952 |
+
"step": 19800
|
| 2953 |
+
},
|
| 2954 |
+
{
|
| 2955 |
+
"epoch": 2.1401617250673857,
|
| 2956 |
+
"grad_norm": 0.6034201383590698,
|
| 2957 |
+
"learning_rate": 0.0004721381543443065,
|
| 2958 |
+
"loss": 3.5728,
|
| 2959 |
+
"step": 19850
|
| 2960 |
+
},
|
| 2961 |
+
{
|
| 2962 |
+
"epoch": 2.1455525606469004,
|
| 2963 |
+
"grad_norm": 0.6693742871284485,
|
| 2964 |
+
"learning_rate": 0.000471814355099838,
|
| 2965 |
+
"loss": 3.5634,
|
| 2966 |
+
"step": 19900
|
| 2967 |
+
},
|
| 2968 |
+
{
|
| 2969 |
+
"epoch": 2.150943396226415,
|
| 2970 |
+
"grad_norm": 0.5606766939163208,
|
| 2971 |
+
"learning_rate": 0.0004714905558553697,
|
| 2972 |
+
"loss": 3.5611,
|
| 2973 |
+
"step": 19950
|
| 2974 |
+
},
|
| 2975 |
+
{
|
| 2976 |
+
"epoch": 2.1563342318059298,
|
| 2977 |
+
"grad_norm": 0.600206196308136,
|
| 2978 |
+
"learning_rate": 0.0004711667566109012,
|
| 2979 |
+
"loss": 3.5606,
|
| 2980 |
+
"step": 20000
|
| 2981 |
+
},
|
| 2982 |
+
{
|
| 2983 |
+
"epoch": 2.1563342318059298,
|
| 2984 |
+
"eval_accuracy": 0.36217603919153124,
|
| 2985 |
+
"eval_loss": 3.5728728771209717,
|
| 2986 |
+
"eval_runtime": 181.6212,
|
| 2987 |
+
"eval_samples_per_second": 99.168,
|
| 2988 |
+
"eval_steps_per_second": 6.2,
|
| 2989 |
+
"step": 20000
|
| 2990 |
+
}
|
| 2991 |
+
],
|
| 2992 |
+
"logging_steps": 50,
|
| 2993 |
+
"max_steps": 92750,
|
| 2994 |
+
"num_input_tokens_seen": 0,
|
| 2995 |
+
"num_train_epochs": 10,
|
| 2996 |
+
"save_steps": 10000,
|
| 2997 |
+
"stateful_callbacks": {
|
| 2998 |
+
"TrainerControl": {
|
| 2999 |
+
"args": {
|
| 3000 |
+
"should_epoch_stop": false,
|
| 3001 |
+
"should_evaluate": false,
|
| 3002 |
+
"should_log": false,
|
| 3003 |
+
"should_save": true,
|
| 3004 |
+
"should_training_stop": false
|
| 3005 |
+
},
|
| 3006 |
+
"attributes": {}
|
| 3007 |
+
}
|
| 3008 |
+
},
|
| 3009 |
+
"total_flos": 1.672138358784e+17,
|
| 3010 |
+
"train_batch_size": 32,
|
| 3011 |
+
"trial_name": null,
|
| 3012 |
+
"trial_params": null
|
| 3013 |
+
}
|
checkpoint-20000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:869e284510a25da056f1bb9a81ca3fecc0da8e7b5155c416dd9c5c0c11ed1983
|
| 3 |
+
size 5304
|