Upload folder using huggingface_hub
Browse files- added_tokens.json +3 -0
- config.json +37 -0
- generation_config.json +13 -0
- model.safetensors +3 -0
- optimizer.pt +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- special_tokens_map.json +27 -0
- tokenizer.model +3 -0
- tokenizer_config.json +0 -0
- trainer_state.json +2428 -0
- training_args.bin +3 -0
added_tokens.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"<image_soft_token>": 262144
|
| 3 |
+
}
|
config.json
ADDED
|
@@ -0,0 +1,37 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"Gemma3ForCausalLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_bias": false,
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"attn_logit_softcapping": null,
|
| 8 |
+
"bos_token_id": 2,
|
| 9 |
+
"cache_implementation": "hybrid",
|
| 10 |
+
"eos_token_id": [
|
| 11 |
+
1,
|
| 12 |
+
106
|
| 13 |
+
],
|
| 14 |
+
"final_logit_softcapping": null,
|
| 15 |
+
"head_dim": 256,
|
| 16 |
+
"hidden_activation": "gelu_pytorch_tanh",
|
| 17 |
+
"hidden_size": 1152,
|
| 18 |
+
"initializer_range": 0.02,
|
| 19 |
+
"intermediate_size": 6912,
|
| 20 |
+
"max_position_embeddings": 32768,
|
| 21 |
+
"model_type": "gemma3_text",
|
| 22 |
+
"num_attention_heads": 4,
|
| 23 |
+
"num_hidden_layers": 26,
|
| 24 |
+
"num_key_value_heads": 1,
|
| 25 |
+
"pad_token_id": 0,
|
| 26 |
+
"query_pre_attn_scalar": 256,
|
| 27 |
+
"rms_norm_eps": 1e-06,
|
| 28 |
+
"rope_local_base_freq": 10000,
|
| 29 |
+
"rope_scaling": null,
|
| 30 |
+
"rope_theta": 1000000,
|
| 31 |
+
"sliding_window": 512,
|
| 32 |
+
"sliding_window_pattern": 6,
|
| 33 |
+
"torch_dtype": "bfloat16",
|
| 34 |
+
"transformers_version": "4.51.3",
|
| 35 |
+
"use_cache": true,
|
| 36 |
+
"vocab_size": 262144
|
| 37 |
+
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token_id": 2,
|
| 3 |
+
"cache_implementation": "hybrid",
|
| 4 |
+
"do_sample": true,
|
| 5 |
+
"eos_token_id": [
|
| 6 |
+
1,
|
| 7 |
+
106
|
| 8 |
+
],
|
| 9 |
+
"pad_token_id": 0,
|
| 10 |
+
"top_k": 64,
|
| 11 |
+
"top_p": 0.95,
|
| 12 |
+
"transformers_version": "4.51.3"
|
| 13 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9aa3c7cd374f77af925b6a4358dd0f28084d81dc966725d2e11a3d03270a9a19
|
| 3 |
+
size 1999811208
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c7aef3f7bceb332deeded129c32c933c3d084644ef554a9cfa793e0c0fd9b2b3
|
| 3 |
+
size 3999829006
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ea8c241b0faa82f9f811aceb142306562b098f3ad2bbe260a7ecbf6d86b599ba
|
| 3 |
+
size 14244
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a06e7b51fd132a96e59fc8c1ee1ad430398f425a7c607076821686d60122d971
|
| 3 |
+
size 1064
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,27 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"boi_token": "<start_of_image>",
|
| 3 |
+
"bos_token": {
|
| 4 |
+
"content": "<bos>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false
|
| 9 |
+
},
|
| 10 |
+
"eoi_token": "<end_of_image>",
|
| 11 |
+
"eos_token": {
|
| 12 |
+
"content": "<eos>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false
|
| 17 |
+
},
|
| 18 |
+
"image_token": "<image_soft_token>",
|
| 19 |
+
"pad_token": "<eos>",
|
| 20 |
+
"unk_token": {
|
| 21 |
+
"content": "<unk>",
|
| 22 |
+
"lstrip": false,
|
| 23 |
+
"normalized": false,
|
| 24 |
+
"rstrip": false,
|
| 25 |
+
"single_word": false
|
| 26 |
+
}
|
| 27 |
+
}
|
tokenizer.model
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1299c11d7cf632ef3b4e11937501358ada021bbdf7c47638d13c0ee982f2e79c
|
| 3 |
+
size 4689074
|
tokenizer_config.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2428 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 3.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 17121,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.008761170492377781,
|
| 14 |
+
"grad_norm": 1.859375,
|
| 15 |
+
"learning_rate": 4.985690088195783e-05,
|
| 16 |
+
"loss": 0.5628,
|
| 17 |
+
"step": 50
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.017522340984755563,
|
| 21 |
+
"grad_norm": 1.453125,
|
| 22 |
+
"learning_rate": 4.971088137375153e-05,
|
| 23 |
+
"loss": 0.3821,
|
| 24 |
+
"step": 100
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.026283511477133346,
|
| 28 |
+
"grad_norm": 1.2734375,
|
| 29 |
+
"learning_rate": 4.9564861865545236e-05,
|
| 30 |
+
"loss": 0.3687,
|
| 31 |
+
"step": 150
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.035044681969511125,
|
| 35 |
+
"grad_norm": 1.234375,
|
| 36 |
+
"learning_rate": 4.941884235733894e-05,
|
| 37 |
+
"loss": 0.3548,
|
| 38 |
+
"step": 200
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.04380585246188891,
|
| 42 |
+
"grad_norm": 1.0625,
|
| 43 |
+
"learning_rate": 4.927282284913264e-05,
|
| 44 |
+
"loss": 0.3448,
|
| 45 |
+
"step": 250
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.05256702295426669,
|
| 49 |
+
"grad_norm": 1.140625,
|
| 50 |
+
"learning_rate": 4.912680334092635e-05,
|
| 51 |
+
"loss": 0.3381,
|
| 52 |
+
"step": 300
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.061328193446644474,
|
| 56 |
+
"grad_norm": 1.1328125,
|
| 57 |
+
"learning_rate": 4.898078383272006e-05,
|
| 58 |
+
"loss": 0.3325,
|
| 59 |
+
"step": 350
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.07008936393902225,
|
| 63 |
+
"grad_norm": 1.0546875,
|
| 64 |
+
"learning_rate": 4.883476432451376e-05,
|
| 65 |
+
"loss": 0.324,
|
| 66 |
+
"step": 400
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.07885053443140004,
|
| 70 |
+
"grad_norm": 0.9453125,
|
| 71 |
+
"learning_rate": 4.8688744816307465e-05,
|
| 72 |
+
"loss": 0.3223,
|
| 73 |
+
"step": 450
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.08761170492377782,
|
| 77 |
+
"grad_norm": 0.984375,
|
| 78 |
+
"learning_rate": 4.854272530810117e-05,
|
| 79 |
+
"loss": 0.3197,
|
| 80 |
+
"step": 500
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.09637287541615559,
|
| 84 |
+
"grad_norm": 0.94140625,
|
| 85 |
+
"learning_rate": 4.839670579989487e-05,
|
| 86 |
+
"loss": 0.3086,
|
| 87 |
+
"step": 550
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.10513404590853338,
|
| 91 |
+
"grad_norm": 0.9453125,
|
| 92 |
+
"learning_rate": 4.825068629168857e-05,
|
| 93 |
+
"loss": 0.3122,
|
| 94 |
+
"step": 600
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.11389521640091116,
|
| 98 |
+
"grad_norm": 0.90234375,
|
| 99 |
+
"learning_rate": 4.8104666783482275e-05,
|
| 100 |
+
"loss": 0.3035,
|
| 101 |
+
"step": 650
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.12265638689328895,
|
| 105 |
+
"grad_norm": 0.91796875,
|
| 106 |
+
"learning_rate": 4.795864727527598e-05,
|
| 107 |
+
"loss": 0.3073,
|
| 108 |
+
"step": 700
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.13141755738566674,
|
| 112 |
+
"grad_norm": 0.97265625,
|
| 113 |
+
"learning_rate": 4.781262776706969e-05,
|
| 114 |
+
"loss": 0.3051,
|
| 115 |
+
"step": 750
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.1401787278780445,
|
| 119 |
+
"grad_norm": 1.0,
|
| 120 |
+
"learning_rate": 4.766660825886339e-05,
|
| 121 |
+
"loss": 0.2992,
|
| 122 |
+
"step": 800
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.1489398983704223,
|
| 126 |
+
"grad_norm": 0.87890625,
|
| 127 |
+
"learning_rate": 4.752058875065709e-05,
|
| 128 |
+
"loss": 0.3007,
|
| 129 |
+
"step": 850
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.15770106886280008,
|
| 133 |
+
"grad_norm": 0.8828125,
|
| 134 |
+
"learning_rate": 4.7374569242450795e-05,
|
| 135 |
+
"loss": 0.2952,
|
| 136 |
+
"step": 900
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.16646223935517784,
|
| 140 |
+
"grad_norm": 0.875,
|
| 141 |
+
"learning_rate": 4.72285497342445e-05,
|
| 142 |
+
"loss": 0.2981,
|
| 143 |
+
"step": 950
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.17522340984755563,
|
| 147 |
+
"grad_norm": 0.828125,
|
| 148 |
+
"learning_rate": 4.70825302260382e-05,
|
| 149 |
+
"loss": 0.2911,
|
| 150 |
+
"step": 1000
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.18398458033993342,
|
| 154 |
+
"grad_norm": 0.83203125,
|
| 155 |
+
"learning_rate": 4.69365107178319e-05,
|
| 156 |
+
"loss": 0.2848,
|
| 157 |
+
"step": 1050
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.19274575083231119,
|
| 161 |
+
"grad_norm": 0.80078125,
|
| 162 |
+
"learning_rate": 4.679049120962561e-05,
|
| 163 |
+
"loss": 0.2932,
|
| 164 |
+
"step": 1100
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.20150692132468898,
|
| 168 |
+
"grad_norm": 0.828125,
|
| 169 |
+
"learning_rate": 4.6644471701419314e-05,
|
| 170 |
+
"loss": 0.2908,
|
| 171 |
+
"step": 1150
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.21026809181706677,
|
| 175 |
+
"grad_norm": 0.9140625,
|
| 176 |
+
"learning_rate": 4.649845219321302e-05,
|
| 177 |
+
"loss": 0.2836,
|
| 178 |
+
"step": 1200
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.21902926230944456,
|
| 182 |
+
"grad_norm": 0.80859375,
|
| 183 |
+
"learning_rate": 4.635243268500672e-05,
|
| 184 |
+
"loss": 0.2907,
|
| 185 |
+
"step": 1250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.22779043280182232,
|
| 189 |
+
"grad_norm": 0.87109375,
|
| 190 |
+
"learning_rate": 4.620641317680042e-05,
|
| 191 |
+
"loss": 0.2859,
|
| 192 |
+
"step": 1300
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.2365516032942001,
|
| 196 |
+
"grad_norm": 0.85546875,
|
| 197 |
+
"learning_rate": 4.6060393668594125e-05,
|
| 198 |
+
"loss": 0.2826,
|
| 199 |
+
"step": 1350
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.2453127737865779,
|
| 203 |
+
"grad_norm": 0.859375,
|
| 204 |
+
"learning_rate": 4.591437416038783e-05,
|
| 205 |
+
"loss": 0.2862,
|
| 206 |
+
"step": 1400
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.2540739442789557,
|
| 210 |
+
"grad_norm": 0.8203125,
|
| 211 |
+
"learning_rate": 4.576835465218153e-05,
|
| 212 |
+
"loss": 0.2865,
|
| 213 |
+
"step": 1450
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.2628351147713335,
|
| 217 |
+
"grad_norm": 0.8046875,
|
| 218 |
+
"learning_rate": 4.562233514397524e-05,
|
| 219 |
+
"loss": 0.2792,
|
| 220 |
+
"step": 1500
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.2715962852637112,
|
| 224 |
+
"grad_norm": 1.0703125,
|
| 225 |
+
"learning_rate": 4.547631563576894e-05,
|
| 226 |
+
"loss": 0.281,
|
| 227 |
+
"step": 1550
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.280357455756089,
|
| 231 |
+
"grad_norm": 0.91015625,
|
| 232 |
+
"learning_rate": 4.5330296127562645e-05,
|
| 233 |
+
"loss": 0.28,
|
| 234 |
+
"step": 1600
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.2891186262484668,
|
| 238 |
+
"grad_norm": 0.89453125,
|
| 239 |
+
"learning_rate": 4.518427661935635e-05,
|
| 240 |
+
"loss": 0.2789,
|
| 241 |
+
"step": 1650
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.2978797967408446,
|
| 245 |
+
"grad_norm": 0.8984375,
|
| 246 |
+
"learning_rate": 4.503825711115005e-05,
|
| 247 |
+
"loss": 0.2792,
|
| 248 |
+
"step": 1700
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.3066409672332224,
|
| 252 |
+
"grad_norm": 0.765625,
|
| 253 |
+
"learning_rate": 4.489223760294375e-05,
|
| 254 |
+
"loss": 0.2809,
|
| 255 |
+
"step": 1750
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.31540213772560016,
|
| 259 |
+
"grad_norm": 0.859375,
|
| 260 |
+
"learning_rate": 4.4746218094737455e-05,
|
| 261 |
+
"loss": 0.2729,
|
| 262 |
+
"step": 1800
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.3241633082179779,
|
| 266 |
+
"grad_norm": 0.82421875,
|
| 267 |
+
"learning_rate": 4.4600198586531164e-05,
|
| 268 |
+
"loss": 0.2763,
|
| 269 |
+
"step": 1850
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.3329244787103557,
|
| 273 |
+
"grad_norm": 0.83984375,
|
| 274 |
+
"learning_rate": 4.445417907832487e-05,
|
| 275 |
+
"loss": 0.2789,
|
| 276 |
+
"step": 1900
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.3416856492027335,
|
| 280 |
+
"grad_norm": 0.8671875,
|
| 281 |
+
"learning_rate": 4.430815957011857e-05,
|
| 282 |
+
"loss": 0.2775,
|
| 283 |
+
"step": 1950
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.35044681969511127,
|
| 287 |
+
"grad_norm": 0.87890625,
|
| 288 |
+
"learning_rate": 4.416214006191227e-05,
|
| 289 |
+
"loss": 0.2727,
|
| 290 |
+
"step": 2000
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.35920799018748906,
|
| 294 |
+
"grad_norm": 0.87109375,
|
| 295 |
+
"learning_rate": 4.4016120553705975e-05,
|
| 296 |
+
"loss": 0.2748,
|
| 297 |
+
"step": 2050
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.36796916067986685,
|
| 301 |
+
"grad_norm": 0.81640625,
|
| 302 |
+
"learning_rate": 4.387010104549968e-05,
|
| 303 |
+
"loss": 0.2747,
|
| 304 |
+
"step": 2100
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.37673033117224464,
|
| 308 |
+
"grad_norm": 0.875,
|
| 309 |
+
"learning_rate": 4.372408153729338e-05,
|
| 310 |
+
"loss": 0.2711,
|
| 311 |
+
"step": 2150
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.38549150166462237,
|
| 315 |
+
"grad_norm": 0.72265625,
|
| 316 |
+
"learning_rate": 4.357806202908708e-05,
|
| 317 |
+
"loss": 0.2694,
|
| 318 |
+
"step": 2200
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.39425267215700016,
|
| 322 |
+
"grad_norm": 0.80078125,
|
| 323 |
+
"learning_rate": 4.343204252088079e-05,
|
| 324 |
+
"loss": 0.27,
|
| 325 |
+
"step": 2250
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.40301384264937795,
|
| 329 |
+
"grad_norm": 0.76171875,
|
| 330 |
+
"learning_rate": 4.32860230126745e-05,
|
| 331 |
+
"loss": 0.2662,
|
| 332 |
+
"step": 2300
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.41177501314175574,
|
| 336 |
+
"grad_norm": 0.81640625,
|
| 337 |
+
"learning_rate": 4.3140003504468204e-05,
|
| 338 |
+
"loss": 0.2764,
|
| 339 |
+
"step": 2350
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.42053618363413353,
|
| 343 |
+
"grad_norm": 0.84765625,
|
| 344 |
+
"learning_rate": 4.2993983996261906e-05,
|
| 345 |
+
"loss": 0.2709,
|
| 346 |
+
"step": 2400
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.4292973541265113,
|
| 350 |
+
"grad_norm": 0.921875,
|
| 351 |
+
"learning_rate": 4.284796448805561e-05,
|
| 352 |
+
"loss": 0.2694,
|
| 353 |
+
"step": 2450
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.4380585246188891,
|
| 357 |
+
"grad_norm": 0.77734375,
|
| 358 |
+
"learning_rate": 4.270194497984931e-05,
|
| 359 |
+
"loss": 0.2673,
|
| 360 |
+
"step": 2500
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.44681969511126685,
|
| 364 |
+
"grad_norm": 0.78515625,
|
| 365 |
+
"learning_rate": 4.2555925471643014e-05,
|
| 366 |
+
"loss": 0.2669,
|
| 367 |
+
"step": 2550
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.45558086560364464,
|
| 371 |
+
"grad_norm": 0.6796875,
|
| 372 |
+
"learning_rate": 4.240990596343672e-05,
|
| 373 |
+
"loss": 0.2686,
|
| 374 |
+
"step": 2600
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.4643420360960224,
|
| 378 |
+
"grad_norm": 0.765625,
|
| 379 |
+
"learning_rate": 4.226388645523042e-05,
|
| 380 |
+
"loss": 0.2663,
|
| 381 |
+
"step": 2650
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.4731032065884002,
|
| 385 |
+
"grad_norm": 0.7734375,
|
| 386 |
+
"learning_rate": 4.211786694702413e-05,
|
| 387 |
+
"loss": 0.2638,
|
| 388 |
+
"step": 2700
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.481864377080778,
|
| 392 |
+
"grad_norm": 0.765625,
|
| 393 |
+
"learning_rate": 4.197184743881783e-05,
|
| 394 |
+
"loss": 0.2644,
|
| 395 |
+
"step": 2750
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.4906255475731558,
|
| 399 |
+
"grad_norm": 0.921875,
|
| 400 |
+
"learning_rate": 4.1825827930611534e-05,
|
| 401 |
+
"loss": 0.2643,
|
| 402 |
+
"step": 2800
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.49938671806553353,
|
| 406 |
+
"grad_norm": 0.7421875,
|
| 407 |
+
"learning_rate": 4.1679808422405236e-05,
|
| 408 |
+
"loss": 0.2675,
|
| 409 |
+
"step": 2850
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.5081478885579114,
|
| 413 |
+
"grad_norm": 0.9140625,
|
| 414 |
+
"learning_rate": 4.153378891419894e-05,
|
| 415 |
+
"loss": 0.269,
|
| 416 |
+
"step": 2900
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.5169090590502892,
|
| 420 |
+
"grad_norm": 0.84765625,
|
| 421 |
+
"learning_rate": 4.138776940599264e-05,
|
| 422 |
+
"loss": 0.2607,
|
| 423 |
+
"step": 2950
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.525670229542667,
|
| 427 |
+
"grad_norm": 0.83203125,
|
| 428 |
+
"learning_rate": 4.1241749897786344e-05,
|
| 429 |
+
"loss": 0.2598,
|
| 430 |
+
"step": 3000
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.5344314000350446,
|
| 434 |
+
"grad_norm": 0.75390625,
|
| 435 |
+
"learning_rate": 4.1095730389580054e-05,
|
| 436 |
+
"loss": 0.2649,
|
| 437 |
+
"step": 3050
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.5431925705274224,
|
| 441 |
+
"grad_norm": 0.71484375,
|
| 442 |
+
"learning_rate": 4.0949710881373756e-05,
|
| 443 |
+
"loss": 0.2636,
|
| 444 |
+
"step": 3100
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.5519537410198002,
|
| 448 |
+
"grad_norm": 0.73046875,
|
| 449 |
+
"learning_rate": 4.080369137316746e-05,
|
| 450 |
+
"loss": 0.2603,
|
| 451 |
+
"step": 3150
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.560714911512178,
|
| 455 |
+
"grad_norm": 0.8203125,
|
| 456 |
+
"learning_rate": 4.065767186496116e-05,
|
| 457 |
+
"loss": 0.2566,
|
| 458 |
+
"step": 3200
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.5694760820045558,
|
| 462 |
+
"grad_norm": 0.79296875,
|
| 463 |
+
"learning_rate": 4.0511652356754864e-05,
|
| 464 |
+
"loss": 0.2577,
|
| 465 |
+
"step": 3250
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.5782372524969336,
|
| 469 |
+
"grad_norm": 0.828125,
|
| 470 |
+
"learning_rate": 4.0365632848548566e-05,
|
| 471 |
+
"loss": 0.2596,
|
| 472 |
+
"step": 3300
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.5869984229893114,
|
| 476 |
+
"grad_norm": 0.8125,
|
| 477 |
+
"learning_rate": 4.021961334034227e-05,
|
| 478 |
+
"loss": 0.2622,
|
| 479 |
+
"step": 3350
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.5957595934816892,
|
| 483 |
+
"grad_norm": 0.7734375,
|
| 484 |
+
"learning_rate": 4.007359383213597e-05,
|
| 485 |
+
"loss": 0.2584,
|
| 486 |
+
"step": 3400
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.604520763974067,
|
| 490 |
+
"grad_norm": 0.79296875,
|
| 491 |
+
"learning_rate": 3.992757432392968e-05,
|
| 492 |
+
"loss": 0.2572,
|
| 493 |
+
"step": 3450
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.6132819344664447,
|
| 497 |
+
"grad_norm": 0.84765625,
|
| 498 |
+
"learning_rate": 3.9781554815723384e-05,
|
| 499 |
+
"loss": 0.26,
|
| 500 |
+
"step": 3500
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.6220431049588225,
|
| 504 |
+
"grad_norm": 0.80859375,
|
| 505 |
+
"learning_rate": 3.9635535307517086e-05,
|
| 506 |
+
"loss": 0.26,
|
| 507 |
+
"step": 3550
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.6308042754512003,
|
| 511 |
+
"grad_norm": 0.82421875,
|
| 512 |
+
"learning_rate": 3.948951579931079e-05,
|
| 513 |
+
"loss": 0.2586,
|
| 514 |
+
"step": 3600
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.6395654459435781,
|
| 518 |
+
"grad_norm": 0.8984375,
|
| 519 |
+
"learning_rate": 3.934349629110449e-05,
|
| 520 |
+
"loss": 0.2652,
|
| 521 |
+
"step": 3650
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.6483266164359558,
|
| 525 |
+
"grad_norm": 0.77734375,
|
| 526 |
+
"learning_rate": 3.9197476782898194e-05,
|
| 527 |
+
"loss": 0.2576,
|
| 528 |
+
"step": 3700
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.6570877869283336,
|
| 532 |
+
"grad_norm": 0.7890625,
|
| 533 |
+
"learning_rate": 3.9051457274691897e-05,
|
| 534 |
+
"loss": 0.2595,
|
| 535 |
+
"step": 3750
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.6658489574207114,
|
| 539 |
+
"grad_norm": 0.7890625,
|
| 540 |
+
"learning_rate": 3.8905437766485606e-05,
|
| 541 |
+
"loss": 0.256,
|
| 542 |
+
"step": 3800
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.6746101279130892,
|
| 546 |
+
"grad_norm": 0.84375,
|
| 547 |
+
"learning_rate": 3.875941825827931e-05,
|
| 548 |
+
"loss": 0.2571,
|
| 549 |
+
"step": 3850
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.683371298405467,
|
| 553 |
+
"grad_norm": 0.8671875,
|
| 554 |
+
"learning_rate": 3.861339875007301e-05,
|
| 555 |
+
"loss": 0.2546,
|
| 556 |
+
"step": 3900
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.6921324688978447,
|
| 560 |
+
"grad_norm": 0.875,
|
| 561 |
+
"learning_rate": 3.8467379241866714e-05,
|
| 562 |
+
"loss": 0.2578,
|
| 563 |
+
"step": 3950
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.7008936393902225,
|
| 567 |
+
"grad_norm": 0.81640625,
|
| 568 |
+
"learning_rate": 3.8321359733660416e-05,
|
| 569 |
+
"loss": 0.2551,
|
| 570 |
+
"step": 4000
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.7096548098826003,
|
| 574 |
+
"grad_norm": 0.80859375,
|
| 575 |
+
"learning_rate": 3.817534022545412e-05,
|
| 576 |
+
"loss": 0.2586,
|
| 577 |
+
"step": 4050
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.7184159803749781,
|
| 581 |
+
"grad_norm": 0.8046875,
|
| 582 |
+
"learning_rate": 3.802932071724782e-05,
|
| 583 |
+
"loss": 0.2568,
|
| 584 |
+
"step": 4100
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.7271771508673559,
|
| 588 |
+
"grad_norm": 0.8515625,
|
| 589 |
+
"learning_rate": 3.7883301209041524e-05,
|
| 590 |
+
"loss": 0.2632,
|
| 591 |
+
"step": 4150
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.7359383213597337,
|
| 595 |
+
"grad_norm": 0.75390625,
|
| 596 |
+
"learning_rate": 3.7737281700835233e-05,
|
| 597 |
+
"loss": 0.2518,
|
| 598 |
+
"step": 4200
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.7446994918521115,
|
| 602 |
+
"grad_norm": 0.8203125,
|
| 603 |
+
"learning_rate": 3.7591262192628936e-05,
|
| 604 |
+
"loss": 0.2539,
|
| 605 |
+
"step": 4250
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.7534606623444893,
|
| 609 |
+
"grad_norm": 0.765625,
|
| 610 |
+
"learning_rate": 3.7445242684422645e-05,
|
| 611 |
+
"loss": 0.2507,
|
| 612 |
+
"step": 4300
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.762221832836867,
|
| 616 |
+
"grad_norm": 0.83984375,
|
| 617 |
+
"learning_rate": 3.729922317621635e-05,
|
| 618 |
+
"loss": 0.2517,
|
| 619 |
+
"step": 4350
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.7709830033292447,
|
| 623 |
+
"grad_norm": 0.74609375,
|
| 624 |
+
"learning_rate": 3.715320366801005e-05,
|
| 625 |
+
"loss": 0.2523,
|
| 626 |
+
"step": 4400
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.7797441738216225,
|
| 630 |
+
"grad_norm": 0.70703125,
|
| 631 |
+
"learning_rate": 3.700718415980375e-05,
|
| 632 |
+
"loss": 0.2492,
|
| 633 |
+
"step": 4450
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.7885053443140003,
|
| 637 |
+
"grad_norm": 0.8359375,
|
| 638 |
+
"learning_rate": 3.6861164651597456e-05,
|
| 639 |
+
"loss": 0.2571,
|
| 640 |
+
"step": 4500
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.7972665148063781,
|
| 644 |
+
"grad_norm": 0.7890625,
|
| 645 |
+
"learning_rate": 3.671514514339116e-05,
|
| 646 |
+
"loss": 0.2534,
|
| 647 |
+
"step": 4550
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.8060276852987559,
|
| 651 |
+
"grad_norm": 0.79296875,
|
| 652 |
+
"learning_rate": 3.656912563518487e-05,
|
| 653 |
+
"loss": 0.2573,
|
| 654 |
+
"step": 4600
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.8147888557911337,
|
| 658 |
+
"grad_norm": 0.80078125,
|
| 659 |
+
"learning_rate": 3.642310612697857e-05,
|
| 660 |
+
"loss": 0.2552,
|
| 661 |
+
"step": 4650
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.8235500262835115,
|
| 665 |
+
"grad_norm": 0.796875,
|
| 666 |
+
"learning_rate": 3.627708661877227e-05,
|
| 667 |
+
"loss": 0.252,
|
| 668 |
+
"step": 4700
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.8323111967758893,
|
| 672 |
+
"grad_norm": 0.75,
|
| 673 |
+
"learning_rate": 3.6131067110565975e-05,
|
| 674 |
+
"loss": 0.2545,
|
| 675 |
+
"step": 4750
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.8410723672682671,
|
| 679 |
+
"grad_norm": 0.77734375,
|
| 680 |
+
"learning_rate": 3.598504760235968e-05,
|
| 681 |
+
"loss": 0.2536,
|
| 682 |
+
"step": 4800
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.8498335377606449,
|
| 686 |
+
"grad_norm": 0.83203125,
|
| 687 |
+
"learning_rate": 3.583902809415338e-05,
|
| 688 |
+
"loss": 0.2486,
|
| 689 |
+
"step": 4850
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.8585947082530226,
|
| 693 |
+
"grad_norm": 0.79296875,
|
| 694 |
+
"learning_rate": 3.569300858594708e-05,
|
| 695 |
+
"loss": 0.2491,
|
| 696 |
+
"step": 4900
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.8673558787454004,
|
| 700 |
+
"grad_norm": 0.796875,
|
| 701 |
+
"learning_rate": 3.5546989077740786e-05,
|
| 702 |
+
"loss": 0.2493,
|
| 703 |
+
"step": 4950
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.8761170492377782,
|
| 707 |
+
"grad_norm": 0.8359375,
|
| 708 |
+
"learning_rate": 3.5400969569534495e-05,
|
| 709 |
+
"loss": 0.2496,
|
| 710 |
+
"step": 5000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.8848782197301559,
|
| 714 |
+
"grad_norm": 0.73828125,
|
| 715 |
+
"learning_rate": 3.52549500613282e-05,
|
| 716 |
+
"loss": 0.2488,
|
| 717 |
+
"step": 5050
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.8936393902225337,
|
| 721 |
+
"grad_norm": 0.8046875,
|
| 722 |
+
"learning_rate": 3.51089305531219e-05,
|
| 723 |
+
"loss": 0.2497,
|
| 724 |
+
"step": 5100
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 0.9024005607149115,
|
| 728 |
+
"grad_norm": 0.80078125,
|
| 729 |
+
"learning_rate": 3.49629110449156e-05,
|
| 730 |
+
"loss": 0.253,
|
| 731 |
+
"step": 5150
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.9111617312072893,
|
| 735 |
+
"grad_norm": 0.7421875,
|
| 736 |
+
"learning_rate": 3.4816891536709306e-05,
|
| 737 |
+
"loss": 0.2515,
|
| 738 |
+
"step": 5200
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.9199229016996671,
|
| 742 |
+
"grad_norm": 0.7890625,
|
| 743 |
+
"learning_rate": 3.467087202850301e-05,
|
| 744 |
+
"loss": 0.2479,
|
| 745 |
+
"step": 5250
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.9286840721920449,
|
| 749 |
+
"grad_norm": 0.796875,
|
| 750 |
+
"learning_rate": 3.452485252029671e-05,
|
| 751 |
+
"loss": 0.2486,
|
| 752 |
+
"step": 5300
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.9374452426844226,
|
| 756 |
+
"grad_norm": 0.82421875,
|
| 757 |
+
"learning_rate": 3.437883301209042e-05,
|
| 758 |
+
"loss": 0.2518,
|
| 759 |
+
"step": 5350
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.9462064131768004,
|
| 763 |
+
"grad_norm": 0.85546875,
|
| 764 |
+
"learning_rate": 3.423281350388412e-05,
|
| 765 |
+
"loss": 0.2499,
|
| 766 |
+
"step": 5400
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 0.9549675836691782,
|
| 770 |
+
"grad_norm": 0.76171875,
|
| 771 |
+
"learning_rate": 3.4086793995677825e-05,
|
| 772 |
+
"loss": 0.2508,
|
| 773 |
+
"step": 5450
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.963728754161556,
|
| 777 |
+
"grad_norm": 0.83203125,
|
| 778 |
+
"learning_rate": 3.394077448747153e-05,
|
| 779 |
+
"loss": 0.2482,
|
| 780 |
+
"step": 5500
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 0.9724899246539338,
|
| 784 |
+
"grad_norm": 0.78515625,
|
| 785 |
+
"learning_rate": 3.379475497926523e-05,
|
| 786 |
+
"loss": 0.2491,
|
| 787 |
+
"step": 5550
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 0.9812510951463116,
|
| 791 |
+
"grad_norm": 0.84765625,
|
| 792 |
+
"learning_rate": 3.364873547105893e-05,
|
| 793 |
+
"loss": 0.2471,
|
| 794 |
+
"step": 5600
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.9900122656386894,
|
| 798 |
+
"grad_norm": 0.703125,
|
| 799 |
+
"learning_rate": 3.3502715962852636e-05,
|
| 800 |
+
"loss": 0.2539,
|
| 801 |
+
"step": 5650
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.9987734361310671,
|
| 805 |
+
"grad_norm": 0.77734375,
|
| 806 |
+
"learning_rate": 3.335669645464634e-05,
|
| 807 |
+
"loss": 0.2488,
|
| 808 |
+
"step": 5700
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 1.007534606623445,
|
| 812 |
+
"grad_norm": 0.80078125,
|
| 813 |
+
"learning_rate": 3.321067694644005e-05,
|
| 814 |
+
"loss": 0.2335,
|
| 815 |
+
"step": 5750
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 1.0162957771158228,
|
| 819 |
+
"grad_norm": 0.8046875,
|
| 820 |
+
"learning_rate": 3.306465743823375e-05,
|
| 821 |
+
"loss": 0.2302,
|
| 822 |
+
"step": 5800
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 1.0250569476082005,
|
| 826 |
+
"grad_norm": 0.84765625,
|
| 827 |
+
"learning_rate": 3.291863793002745e-05,
|
| 828 |
+
"loss": 0.2287,
|
| 829 |
+
"step": 5850
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 1.0338181181005783,
|
| 833 |
+
"grad_norm": 0.84765625,
|
| 834 |
+
"learning_rate": 3.2772618421821155e-05,
|
| 835 |
+
"loss": 0.2278,
|
| 836 |
+
"step": 5900
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 1.0425792885929561,
|
| 840 |
+
"grad_norm": 0.87109375,
|
| 841 |
+
"learning_rate": 3.262659891361486e-05,
|
| 842 |
+
"loss": 0.2322,
|
| 843 |
+
"step": 5950
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 1.051340459085334,
|
| 847 |
+
"grad_norm": 0.87109375,
|
| 848 |
+
"learning_rate": 3.248057940540856e-05,
|
| 849 |
+
"loss": 0.2305,
|
| 850 |
+
"step": 6000
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 1.0601016295777115,
|
| 854 |
+
"grad_norm": 0.95703125,
|
| 855 |
+
"learning_rate": 3.233455989720226e-05,
|
| 856 |
+
"loss": 0.2316,
|
| 857 |
+
"step": 6050
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 1.0688628000700893,
|
| 861 |
+
"grad_norm": 0.75,
|
| 862 |
+
"learning_rate": 3.218854038899597e-05,
|
| 863 |
+
"loss": 0.2366,
|
| 864 |
+
"step": 6100
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 1.077623970562467,
|
| 868 |
+
"grad_norm": 1.015625,
|
| 869 |
+
"learning_rate": 3.2042520880789675e-05,
|
| 870 |
+
"loss": 0.2331,
|
| 871 |
+
"step": 6150
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 1.0863851410548448,
|
| 875 |
+
"grad_norm": 0.80078125,
|
| 876 |
+
"learning_rate": 3.189650137258338e-05,
|
| 877 |
+
"loss": 0.2283,
|
| 878 |
+
"step": 6200
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 1.0951463115472226,
|
| 882 |
+
"grad_norm": 0.8828125,
|
| 883 |
+
"learning_rate": 3.175048186437708e-05,
|
| 884 |
+
"loss": 0.2301,
|
| 885 |
+
"step": 6250
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 1.1039074820396004,
|
| 889 |
+
"grad_norm": 0.87109375,
|
| 890 |
+
"learning_rate": 3.160446235617079e-05,
|
| 891 |
+
"loss": 0.2311,
|
| 892 |
+
"step": 6300
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 1.1126686525319782,
|
| 896 |
+
"grad_norm": 0.83203125,
|
| 897 |
+
"learning_rate": 3.145844284796449e-05,
|
| 898 |
+
"loss": 0.2278,
|
| 899 |
+
"step": 6350
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 1.121429823024356,
|
| 903 |
+
"grad_norm": 0.8359375,
|
| 904 |
+
"learning_rate": 3.1312423339758195e-05,
|
| 905 |
+
"loss": 0.231,
|
| 906 |
+
"step": 6400
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 1.1301909935167338,
|
| 910 |
+
"grad_norm": 0.97265625,
|
| 911 |
+
"learning_rate": 3.11664038315519e-05,
|
| 912 |
+
"loss": 0.2302,
|
| 913 |
+
"step": 6450
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 1.1389521640091116,
|
| 917 |
+
"grad_norm": 0.94921875,
|
| 918 |
+
"learning_rate": 3.10203843233456e-05,
|
| 919 |
+
"loss": 0.2311,
|
| 920 |
+
"step": 6500
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 1.1477133345014894,
|
| 924 |
+
"grad_norm": 0.85546875,
|
| 925 |
+
"learning_rate": 3.087436481513931e-05,
|
| 926 |
+
"loss": 0.234,
|
| 927 |
+
"step": 6550
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 1.1564745049938672,
|
| 931 |
+
"grad_norm": 0.8984375,
|
| 932 |
+
"learning_rate": 3.072834530693301e-05,
|
| 933 |
+
"loss": 0.231,
|
| 934 |
+
"step": 6600
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 1.165235675486245,
|
| 938 |
+
"grad_norm": 0.86328125,
|
| 939 |
+
"learning_rate": 3.0582325798726714e-05,
|
| 940 |
+
"loss": 0.2294,
|
| 941 |
+
"step": 6650
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 1.1739968459786227,
|
| 945 |
+
"grad_norm": 0.796875,
|
| 946 |
+
"learning_rate": 3.0436306290520417e-05,
|
| 947 |
+
"loss": 0.2316,
|
| 948 |
+
"step": 6700
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 1.1827580164710005,
|
| 952 |
+
"grad_norm": 0.8984375,
|
| 953 |
+
"learning_rate": 3.029028678231412e-05,
|
| 954 |
+
"loss": 0.2335,
|
| 955 |
+
"step": 6750
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 1.1915191869633783,
|
| 959 |
+
"grad_norm": 1.0234375,
|
| 960 |
+
"learning_rate": 3.0144267274107822e-05,
|
| 961 |
+
"loss": 0.2282,
|
| 962 |
+
"step": 6800
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 1.2002803574557561,
|
| 966 |
+
"grad_norm": 0.85546875,
|
| 967 |
+
"learning_rate": 2.9998247765901528e-05,
|
| 968 |
+
"loss": 0.2327,
|
| 969 |
+
"step": 6850
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 1.209041527948134,
|
| 973 |
+
"grad_norm": 0.8671875,
|
| 974 |
+
"learning_rate": 2.985222825769523e-05,
|
| 975 |
+
"loss": 0.2295,
|
| 976 |
+
"step": 6900
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 1.2178026984405117,
|
| 980 |
+
"grad_norm": 0.85546875,
|
| 981 |
+
"learning_rate": 2.9706208749488933e-05,
|
| 982 |
+
"loss": 0.2306,
|
| 983 |
+
"step": 6950
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 1.2265638689328895,
|
| 987 |
+
"grad_norm": 0.875,
|
| 988 |
+
"learning_rate": 2.9560189241282636e-05,
|
| 989 |
+
"loss": 0.234,
|
| 990 |
+
"step": 7000
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 1.2353250394252673,
|
| 994 |
+
"grad_norm": 0.85546875,
|
| 995 |
+
"learning_rate": 2.9414169733076342e-05,
|
| 996 |
+
"loss": 0.2302,
|
| 997 |
+
"step": 7050
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 1.244086209917645,
|
| 1001 |
+
"grad_norm": 0.9375,
|
| 1002 |
+
"learning_rate": 2.9268150224870045e-05,
|
| 1003 |
+
"loss": 0.2273,
|
| 1004 |
+
"step": 7100
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 1.2528473804100229,
|
| 1008 |
+
"grad_norm": 0.83203125,
|
| 1009 |
+
"learning_rate": 2.9122130716663747e-05,
|
| 1010 |
+
"loss": 0.2255,
|
| 1011 |
+
"step": 7150
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 1.2616085509024004,
|
| 1015 |
+
"grad_norm": 0.8828125,
|
| 1016 |
+
"learning_rate": 2.8976111208457453e-05,
|
| 1017 |
+
"loss": 0.2305,
|
| 1018 |
+
"step": 7200
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 1.2703697213947782,
|
| 1022 |
+
"grad_norm": 0.8125,
|
| 1023 |
+
"learning_rate": 2.8830091700251156e-05,
|
| 1024 |
+
"loss": 0.2293,
|
| 1025 |
+
"step": 7250
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 1.279130891887156,
|
| 1029 |
+
"grad_norm": 0.91796875,
|
| 1030 |
+
"learning_rate": 2.868407219204486e-05,
|
| 1031 |
+
"loss": 0.2258,
|
| 1032 |
+
"step": 7300
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 1.2878920623795338,
|
| 1036 |
+
"grad_norm": 0.859375,
|
| 1037 |
+
"learning_rate": 2.853805268383856e-05,
|
| 1038 |
+
"loss": 0.227,
|
| 1039 |
+
"step": 7350
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 1.2966532328719116,
|
| 1043 |
+
"grad_norm": 0.91015625,
|
| 1044 |
+
"learning_rate": 2.8392033175632267e-05,
|
| 1045 |
+
"loss": 0.2292,
|
| 1046 |
+
"step": 7400
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 1.3054144033642894,
|
| 1050 |
+
"grad_norm": 0.8359375,
|
| 1051 |
+
"learning_rate": 2.824601366742597e-05,
|
| 1052 |
+
"loss": 0.228,
|
| 1053 |
+
"step": 7450
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 1.3141755738566672,
|
| 1057 |
+
"grad_norm": 0.859375,
|
| 1058 |
+
"learning_rate": 2.8099994159219672e-05,
|
| 1059 |
+
"loss": 0.2299,
|
| 1060 |
+
"step": 7500
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 1.322936744349045,
|
| 1064 |
+
"grad_norm": 0.9453125,
|
| 1065 |
+
"learning_rate": 2.7953974651013375e-05,
|
| 1066 |
+
"loss": 0.225,
|
| 1067 |
+
"step": 7550
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 1.3316979148414227,
|
| 1071 |
+
"grad_norm": 0.82421875,
|
| 1072 |
+
"learning_rate": 2.780795514280708e-05,
|
| 1073 |
+
"loss": 0.2274,
|
| 1074 |
+
"step": 7600
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 1.3404590853338005,
|
| 1078 |
+
"grad_norm": 0.921875,
|
| 1079 |
+
"learning_rate": 2.7661935634600783e-05,
|
| 1080 |
+
"loss": 0.2295,
|
| 1081 |
+
"step": 7650
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 1.3492202558261783,
|
| 1085 |
+
"grad_norm": 0.8359375,
|
| 1086 |
+
"learning_rate": 2.7515916126394486e-05,
|
| 1087 |
+
"loss": 0.2259,
|
| 1088 |
+
"step": 7700
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 1.3579814263185561,
|
| 1092 |
+
"grad_norm": 0.90234375,
|
| 1093 |
+
"learning_rate": 2.736989661818819e-05,
|
| 1094 |
+
"loss": 0.2321,
|
| 1095 |
+
"step": 7750
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 1.366742596810934,
|
| 1099 |
+
"grad_norm": 0.94140625,
|
| 1100 |
+
"learning_rate": 2.7223877109981894e-05,
|
| 1101 |
+
"loss": 0.2279,
|
| 1102 |
+
"step": 7800
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 1.3755037673033117,
|
| 1106 |
+
"grad_norm": 0.8984375,
|
| 1107 |
+
"learning_rate": 2.7077857601775597e-05,
|
| 1108 |
+
"loss": 0.2291,
|
| 1109 |
+
"step": 7850
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 1.3842649377956895,
|
| 1113 |
+
"grad_norm": 0.96484375,
|
| 1114 |
+
"learning_rate": 2.69318380935693e-05,
|
| 1115 |
+
"loss": 0.2273,
|
| 1116 |
+
"step": 7900
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 1.3930261082880673,
|
| 1120 |
+
"grad_norm": 0.9375,
|
| 1121 |
+
"learning_rate": 2.6785818585363006e-05,
|
| 1122 |
+
"loss": 0.2296,
|
| 1123 |
+
"step": 7950
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 1.401787278780445,
|
| 1127 |
+
"grad_norm": 0.91796875,
|
| 1128 |
+
"learning_rate": 2.6639799077156708e-05,
|
| 1129 |
+
"loss": 0.2262,
|
| 1130 |
+
"step": 8000
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 1.4105484492728229,
|
| 1134 |
+
"grad_norm": 0.875,
|
| 1135 |
+
"learning_rate": 2.649377956895041e-05,
|
| 1136 |
+
"loss": 0.2258,
|
| 1137 |
+
"step": 8050
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 1.4193096197652006,
|
| 1141 |
+
"grad_norm": 1.03125,
|
| 1142 |
+
"learning_rate": 2.6347760060744113e-05,
|
| 1143 |
+
"loss": 0.2237,
|
| 1144 |
+
"step": 8100
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 1.4280707902575784,
|
| 1148 |
+
"grad_norm": 0.92578125,
|
| 1149 |
+
"learning_rate": 2.620174055253782e-05,
|
| 1150 |
+
"loss": 0.2306,
|
| 1151 |
+
"step": 8150
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 1.4368319607499562,
|
| 1155 |
+
"grad_norm": 0.95703125,
|
| 1156 |
+
"learning_rate": 2.6055721044331522e-05,
|
| 1157 |
+
"loss": 0.2282,
|
| 1158 |
+
"step": 8200
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 1.445593131242334,
|
| 1162 |
+
"grad_norm": 0.82421875,
|
| 1163 |
+
"learning_rate": 2.5909701536125224e-05,
|
| 1164 |
+
"loss": 0.2245,
|
| 1165 |
+
"step": 8250
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 1.4543543017347118,
|
| 1169 |
+
"grad_norm": 0.890625,
|
| 1170 |
+
"learning_rate": 2.5763682027918934e-05,
|
| 1171 |
+
"loss": 0.2267,
|
| 1172 |
+
"step": 8300
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 1.4631154722270896,
|
| 1176 |
+
"grad_norm": 0.88671875,
|
| 1177 |
+
"learning_rate": 2.5617662519712636e-05,
|
| 1178 |
+
"loss": 0.2248,
|
| 1179 |
+
"step": 8350
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 1.4718766427194674,
|
| 1183 |
+
"grad_norm": 1.109375,
|
| 1184 |
+
"learning_rate": 2.5471643011506342e-05,
|
| 1185 |
+
"loss": 0.2299,
|
| 1186 |
+
"step": 8400
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 1.4806378132118452,
|
| 1190 |
+
"grad_norm": 0.90234375,
|
| 1191 |
+
"learning_rate": 2.5325623503300045e-05,
|
| 1192 |
+
"loss": 0.2222,
|
| 1193 |
+
"step": 8450
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 1.489398983704223,
|
| 1197 |
+
"grad_norm": 0.94921875,
|
| 1198 |
+
"learning_rate": 2.5179603995093748e-05,
|
| 1199 |
+
"loss": 0.228,
|
| 1200 |
+
"step": 8500
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 1.4981601541966008,
|
| 1204 |
+
"grad_norm": 0.83984375,
|
| 1205 |
+
"learning_rate": 2.503358448688745e-05,
|
| 1206 |
+
"loss": 0.2302,
|
| 1207 |
+
"step": 8550
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 1.5069213246889785,
|
| 1211 |
+
"grad_norm": 0.94140625,
|
| 1212 |
+
"learning_rate": 2.4887564978681153e-05,
|
| 1213 |
+
"loss": 0.2281,
|
| 1214 |
+
"step": 8600
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 1.5156824951813563,
|
| 1218 |
+
"grad_norm": 0.88671875,
|
| 1219 |
+
"learning_rate": 2.4741545470474855e-05,
|
| 1220 |
+
"loss": 0.225,
|
| 1221 |
+
"step": 8650
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 1.5244436656737341,
|
| 1225 |
+
"grad_norm": 0.921875,
|
| 1226 |
+
"learning_rate": 2.4595525962268558e-05,
|
| 1227 |
+
"loss": 0.2277,
|
| 1228 |
+
"step": 8700
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 1.533204836166112,
|
| 1232 |
+
"grad_norm": 0.796875,
|
| 1233 |
+
"learning_rate": 2.4449506454062264e-05,
|
| 1234 |
+
"loss": 0.2216,
|
| 1235 |
+
"step": 8750
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 1.5419660066584897,
|
| 1239 |
+
"grad_norm": 0.86328125,
|
| 1240 |
+
"learning_rate": 2.430348694585597e-05,
|
| 1241 |
+
"loss": 0.2238,
|
| 1242 |
+
"step": 8800
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 1.5507271771508675,
|
| 1246 |
+
"grad_norm": 0.91015625,
|
| 1247 |
+
"learning_rate": 2.4157467437649672e-05,
|
| 1248 |
+
"loss": 0.229,
|
| 1249 |
+
"step": 8850
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 1.5594883476432453,
|
| 1253 |
+
"grad_norm": 0.99609375,
|
| 1254 |
+
"learning_rate": 2.4011447929443375e-05,
|
| 1255 |
+
"loss": 0.2273,
|
| 1256 |
+
"step": 8900
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 1.568249518135623,
|
| 1260 |
+
"grad_norm": 0.91796875,
|
| 1261 |
+
"learning_rate": 2.386542842123708e-05,
|
| 1262 |
+
"loss": 0.2261,
|
| 1263 |
+
"step": 8950
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 1.5770106886280009,
|
| 1267 |
+
"grad_norm": 0.921875,
|
| 1268 |
+
"learning_rate": 2.3719408913030784e-05,
|
| 1269 |
+
"loss": 0.2244,
|
| 1270 |
+
"step": 9000
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 1.5857718591203784,
|
| 1274 |
+
"grad_norm": 0.9140625,
|
| 1275 |
+
"learning_rate": 2.3573389404824486e-05,
|
| 1276 |
+
"loss": 0.2243,
|
| 1277 |
+
"step": 9050
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 1.5945330296127562,
|
| 1281 |
+
"grad_norm": 0.93359375,
|
| 1282 |
+
"learning_rate": 2.342736989661819e-05,
|
| 1283 |
+
"loss": 0.2277,
|
| 1284 |
+
"step": 9100
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 1.603294200105134,
|
| 1288 |
+
"grad_norm": 0.83984375,
|
| 1289 |
+
"learning_rate": 2.3281350388411895e-05,
|
| 1290 |
+
"loss": 0.2238,
|
| 1291 |
+
"step": 9150
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 1.6120553705975118,
|
| 1295 |
+
"grad_norm": 0.80078125,
|
| 1296 |
+
"learning_rate": 2.3135330880205597e-05,
|
| 1297 |
+
"loss": 0.2251,
|
| 1298 |
+
"step": 9200
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 1.6208165410898896,
|
| 1302 |
+
"grad_norm": 1.015625,
|
| 1303 |
+
"learning_rate": 2.29893113719993e-05,
|
| 1304 |
+
"loss": 0.2257,
|
| 1305 |
+
"step": 9250
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 1.6295777115822674,
|
| 1309 |
+
"grad_norm": 0.890625,
|
| 1310 |
+
"learning_rate": 2.2843291863793003e-05,
|
| 1311 |
+
"loss": 0.2256,
|
| 1312 |
+
"step": 9300
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 1.6383388820746452,
|
| 1316 |
+
"grad_norm": 1.0078125,
|
| 1317 |
+
"learning_rate": 2.269727235558671e-05,
|
| 1318 |
+
"loss": 0.2239,
|
| 1319 |
+
"step": 9350
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 1.647100052567023,
|
| 1323 |
+
"grad_norm": 0.9140625,
|
| 1324 |
+
"learning_rate": 2.255125284738041e-05,
|
| 1325 |
+
"loss": 0.228,
|
| 1326 |
+
"step": 9400
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 1.6558612230594008,
|
| 1330 |
+
"grad_norm": 1.0546875,
|
| 1331 |
+
"learning_rate": 2.2405233339174114e-05,
|
| 1332 |
+
"loss": 0.2274,
|
| 1333 |
+
"step": 9450
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 1.6646223935517785,
|
| 1337 |
+
"grad_norm": 0.91796875,
|
| 1338 |
+
"learning_rate": 2.2259213830967816e-05,
|
| 1339 |
+
"loss": 0.2214,
|
| 1340 |
+
"step": 9500
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 1.6733835640441563,
|
| 1344 |
+
"grad_norm": 0.9140625,
|
| 1345 |
+
"learning_rate": 2.2113194322761522e-05,
|
| 1346 |
+
"loss": 0.2236,
|
| 1347 |
+
"step": 9550
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 1.6821447345365341,
|
| 1351 |
+
"grad_norm": 0.90625,
|
| 1352 |
+
"learning_rate": 2.1967174814555225e-05,
|
| 1353 |
+
"loss": 0.2222,
|
| 1354 |
+
"step": 9600
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 1.690905905028912,
|
| 1358 |
+
"grad_norm": 0.85546875,
|
| 1359 |
+
"learning_rate": 2.1821155306348927e-05,
|
| 1360 |
+
"loss": 0.2227,
|
| 1361 |
+
"step": 9650
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 1.6996670755212895,
|
| 1365 |
+
"grad_norm": 0.98828125,
|
| 1366 |
+
"learning_rate": 2.1675135798142633e-05,
|
| 1367 |
+
"loss": 0.2277,
|
| 1368 |
+
"step": 9700
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 1.7084282460136673,
|
| 1372 |
+
"grad_norm": 0.9296875,
|
| 1373 |
+
"learning_rate": 2.1529116289936336e-05,
|
| 1374 |
+
"loss": 0.2257,
|
| 1375 |
+
"step": 9750
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 1.717189416506045,
|
| 1379 |
+
"grad_norm": 0.97265625,
|
| 1380 |
+
"learning_rate": 2.1383096781730042e-05,
|
| 1381 |
+
"loss": 0.221,
|
| 1382 |
+
"step": 9800
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 1.7259505869984229,
|
| 1386 |
+
"grad_norm": 0.83203125,
|
| 1387 |
+
"learning_rate": 2.1237077273523745e-05,
|
| 1388 |
+
"loss": 0.2231,
|
| 1389 |
+
"step": 9850
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 1.7347117574908006,
|
| 1393 |
+
"grad_norm": 0.9140625,
|
| 1394 |
+
"learning_rate": 2.1091057765317447e-05,
|
| 1395 |
+
"loss": 0.22,
|
| 1396 |
+
"step": 9900
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 1.7434729279831784,
|
| 1400 |
+
"grad_norm": 0.875,
|
| 1401 |
+
"learning_rate": 2.0945038257111153e-05,
|
| 1402 |
+
"loss": 0.2241,
|
| 1403 |
+
"step": 9950
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 1.7522340984755562,
|
| 1407 |
+
"grad_norm": 0.82421875,
|
| 1408 |
+
"learning_rate": 2.0799018748904856e-05,
|
| 1409 |
+
"loss": 0.2215,
|
| 1410 |
+
"step": 10000
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 1.760995268967934,
|
| 1414 |
+
"grad_norm": 0.828125,
|
| 1415 |
+
"learning_rate": 2.065299924069856e-05,
|
| 1416 |
+
"loss": 0.2267,
|
| 1417 |
+
"step": 10050
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 1.7697564394603118,
|
| 1421 |
+
"grad_norm": 0.9453125,
|
| 1422 |
+
"learning_rate": 2.050697973249226e-05,
|
| 1423 |
+
"loss": 0.222,
|
| 1424 |
+
"step": 10100
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 1.7785176099526896,
|
| 1428 |
+
"grad_norm": 0.92578125,
|
| 1429 |
+
"learning_rate": 2.0360960224285967e-05,
|
| 1430 |
+
"loss": 0.2239,
|
| 1431 |
+
"step": 10150
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 1.7872787804450674,
|
| 1435 |
+
"grad_norm": 0.9140625,
|
| 1436 |
+
"learning_rate": 2.021494071607967e-05,
|
| 1437 |
+
"loss": 0.22,
|
| 1438 |
+
"step": 10200
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 1.7960399509374452,
|
| 1442 |
+
"grad_norm": 0.94140625,
|
| 1443 |
+
"learning_rate": 2.0068921207873372e-05,
|
| 1444 |
+
"loss": 0.2293,
|
| 1445 |
+
"step": 10250
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 1.804801121429823,
|
| 1449 |
+
"grad_norm": 0.96484375,
|
| 1450 |
+
"learning_rate": 1.9922901699667078e-05,
|
| 1451 |
+
"loss": 0.2271,
|
| 1452 |
+
"step": 10300
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 1.8135622919222008,
|
| 1456 |
+
"grad_norm": 0.875,
|
| 1457 |
+
"learning_rate": 1.977688219146078e-05,
|
| 1458 |
+
"loss": 0.2252,
|
| 1459 |
+
"step": 10350
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 1.8223234624145785,
|
| 1463 |
+
"grad_norm": 0.91796875,
|
| 1464 |
+
"learning_rate": 1.9630862683254483e-05,
|
| 1465 |
+
"loss": 0.2265,
|
| 1466 |
+
"step": 10400
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 1.8310846329069563,
|
| 1470 |
+
"grad_norm": 0.9453125,
|
| 1471 |
+
"learning_rate": 1.9484843175048186e-05,
|
| 1472 |
+
"loss": 0.2239,
|
| 1473 |
+
"step": 10450
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 1.8398458033993341,
|
| 1477 |
+
"grad_norm": 0.94921875,
|
| 1478 |
+
"learning_rate": 1.9338823666841892e-05,
|
| 1479 |
+
"loss": 0.222,
|
| 1480 |
+
"step": 10500
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 1.848606973891712,
|
| 1484 |
+
"grad_norm": 1.078125,
|
| 1485 |
+
"learning_rate": 1.9192804158635594e-05,
|
| 1486 |
+
"loss": 0.2231,
|
| 1487 |
+
"step": 10550
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 1.8573681443840897,
|
| 1491 |
+
"grad_norm": 0.98046875,
|
| 1492 |
+
"learning_rate": 1.9046784650429297e-05,
|
| 1493 |
+
"loss": 0.2266,
|
| 1494 |
+
"step": 10600
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 1.8661293148764675,
|
| 1498 |
+
"grad_norm": 0.91015625,
|
| 1499 |
+
"learning_rate": 1.8900765142223e-05,
|
| 1500 |
+
"loss": 0.2245,
|
| 1501 |
+
"step": 10650
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 1.8748904853688453,
|
| 1505 |
+
"grad_norm": 0.98828125,
|
| 1506 |
+
"learning_rate": 1.8754745634016706e-05,
|
| 1507 |
+
"loss": 0.2243,
|
| 1508 |
+
"step": 10700
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 1.883651655861223,
|
| 1512 |
+
"grad_norm": 0.80859375,
|
| 1513 |
+
"learning_rate": 1.8608726125810408e-05,
|
| 1514 |
+
"loss": 0.2194,
|
| 1515 |
+
"step": 10750
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 1.8924128263536009,
|
| 1519 |
+
"grad_norm": 0.87109375,
|
| 1520 |
+
"learning_rate": 1.8462706617604114e-05,
|
| 1521 |
+
"loss": 0.2253,
|
| 1522 |
+
"step": 10800
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 1.9011739968459787,
|
| 1526 |
+
"grad_norm": 0.875,
|
| 1527 |
+
"learning_rate": 1.8316687109397817e-05,
|
| 1528 |
+
"loss": 0.2231,
|
| 1529 |
+
"step": 10850
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 1.9099351673383564,
|
| 1533 |
+
"grad_norm": 0.91015625,
|
| 1534 |
+
"learning_rate": 1.8170667601191523e-05,
|
| 1535 |
+
"loss": 0.2215,
|
| 1536 |
+
"step": 10900
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 1.9186963378307342,
|
| 1540 |
+
"grad_norm": 0.99609375,
|
| 1541 |
+
"learning_rate": 1.8024648092985225e-05,
|
| 1542 |
+
"loss": 0.2246,
|
| 1543 |
+
"step": 10950
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 1.927457508323112,
|
| 1547 |
+
"grad_norm": 0.890625,
|
| 1548 |
+
"learning_rate": 1.7878628584778928e-05,
|
| 1549 |
+
"loss": 0.2215,
|
| 1550 |
+
"step": 11000
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 1.9362186788154898,
|
| 1554 |
+
"grad_norm": 0.953125,
|
| 1555 |
+
"learning_rate": 1.773260907657263e-05,
|
| 1556 |
+
"loss": 0.2251,
|
| 1557 |
+
"step": 11050
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 1.9449798493078676,
|
| 1561 |
+
"grad_norm": 0.91015625,
|
| 1562 |
+
"learning_rate": 1.7586589568366336e-05,
|
| 1563 |
+
"loss": 0.219,
|
| 1564 |
+
"step": 11100
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 1.9537410198002454,
|
| 1568 |
+
"grad_norm": 1.0234375,
|
| 1569 |
+
"learning_rate": 1.744057006016004e-05,
|
| 1570 |
+
"loss": 0.2194,
|
| 1571 |
+
"step": 11150
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 1.9625021902926232,
|
| 1575 |
+
"grad_norm": 0.84765625,
|
| 1576 |
+
"learning_rate": 1.729455055195374e-05,
|
| 1577 |
+
"loss": 0.219,
|
| 1578 |
+
"step": 11200
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 1.971263360785001,
|
| 1582 |
+
"grad_norm": 0.94140625,
|
| 1583 |
+
"learning_rate": 1.7148531043747444e-05,
|
| 1584 |
+
"loss": 0.2183,
|
| 1585 |
+
"step": 11250
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 1.9800245312773788,
|
| 1589 |
+
"grad_norm": 0.8671875,
|
| 1590 |
+
"learning_rate": 1.700251153554115e-05,
|
| 1591 |
+
"loss": 0.221,
|
| 1592 |
+
"step": 11300
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 1.9887857017697566,
|
| 1596 |
+
"grad_norm": 0.921875,
|
| 1597 |
+
"learning_rate": 1.6856492027334853e-05,
|
| 1598 |
+
"loss": 0.2218,
|
| 1599 |
+
"step": 11350
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 1.9975468722621343,
|
| 1603 |
+
"grad_norm": 0.93359375,
|
| 1604 |
+
"learning_rate": 1.6710472519128555e-05,
|
| 1605 |
+
"loss": 0.221,
|
| 1606 |
+
"step": 11400
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 2.006308042754512,
|
| 1610 |
+
"grad_norm": 1.03125,
|
| 1611 |
+
"learning_rate": 1.6564453010922258e-05,
|
| 1612 |
+
"loss": 0.2109,
|
| 1613 |
+
"step": 11450
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 2.01506921324689,
|
| 1617 |
+
"grad_norm": 1.0234375,
|
| 1618 |
+
"learning_rate": 1.6418433502715964e-05,
|
| 1619 |
+
"loss": 0.2091,
|
| 1620 |
+
"step": 11500
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 2.0238303837392677,
|
| 1624 |
+
"grad_norm": 1.078125,
|
| 1625 |
+
"learning_rate": 1.6272413994509666e-05,
|
| 1626 |
+
"loss": 0.2039,
|
| 1627 |
+
"step": 11550
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 2.0325915542316455,
|
| 1631 |
+
"grad_norm": 1.0625,
|
| 1632 |
+
"learning_rate": 1.612639448630337e-05,
|
| 1633 |
+
"loss": 0.2078,
|
| 1634 |
+
"step": 11600
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 2.0413527247240233,
|
| 1638 |
+
"grad_norm": 1.046875,
|
| 1639 |
+
"learning_rate": 1.5980374978097075e-05,
|
| 1640 |
+
"loss": 0.2071,
|
| 1641 |
+
"step": 11650
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 2.050113895216401,
|
| 1645 |
+
"grad_norm": 1.0703125,
|
| 1646 |
+
"learning_rate": 1.5834355469890778e-05,
|
| 1647 |
+
"loss": 0.2056,
|
| 1648 |
+
"step": 11700
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 2.058875065708779,
|
| 1652 |
+
"grad_norm": 1.1015625,
|
| 1653 |
+
"learning_rate": 1.568833596168448e-05,
|
| 1654 |
+
"loss": 0.2097,
|
| 1655 |
+
"step": 11750
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 2.0676362362011567,
|
| 1659 |
+
"grad_norm": 1.109375,
|
| 1660 |
+
"learning_rate": 1.5542316453478186e-05,
|
| 1661 |
+
"loss": 0.2083,
|
| 1662 |
+
"step": 11800
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 2.0763974066935345,
|
| 1666 |
+
"grad_norm": 0.984375,
|
| 1667 |
+
"learning_rate": 1.539629694527189e-05,
|
| 1668 |
+
"loss": 0.2047,
|
| 1669 |
+
"step": 11850
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 2.0851585771859122,
|
| 1673 |
+
"grad_norm": 1.0390625,
|
| 1674 |
+
"learning_rate": 1.5250277437065593e-05,
|
| 1675 |
+
"loss": 0.2044,
|
| 1676 |
+
"step": 11900
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 2.09391974767829,
|
| 1680 |
+
"grad_norm": 1.03125,
|
| 1681 |
+
"learning_rate": 1.5104257928859297e-05,
|
| 1682 |
+
"loss": 0.207,
|
| 1683 |
+
"step": 11950
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 2.102680918170668,
|
| 1687 |
+
"grad_norm": 1.1015625,
|
| 1688 |
+
"learning_rate": 1.4958238420653e-05,
|
| 1689 |
+
"loss": 0.2058,
|
| 1690 |
+
"step": 12000
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 2.111442088663045,
|
| 1694 |
+
"grad_norm": 1.09375,
|
| 1695 |
+
"learning_rate": 1.4812218912446704e-05,
|
| 1696 |
+
"loss": 0.2035,
|
| 1697 |
+
"step": 12050
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 2.120203259155423,
|
| 1701 |
+
"grad_norm": 1.125,
|
| 1702 |
+
"learning_rate": 1.4666199404240409e-05,
|
| 1703 |
+
"loss": 0.21,
|
| 1704 |
+
"step": 12100
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 2.1289644296478007,
|
| 1708 |
+
"grad_norm": 1.015625,
|
| 1709 |
+
"learning_rate": 1.4520179896034111e-05,
|
| 1710 |
+
"loss": 0.2079,
|
| 1711 |
+
"step": 12150
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 2.1377256001401785,
|
| 1715 |
+
"grad_norm": 1.015625,
|
| 1716 |
+
"learning_rate": 1.4374160387827815e-05,
|
| 1717 |
+
"loss": 0.2057,
|
| 1718 |
+
"step": 12200
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 2.1464867706325563,
|
| 1722 |
+
"grad_norm": 1.046875,
|
| 1723 |
+
"learning_rate": 1.4228140879621518e-05,
|
| 1724 |
+
"loss": 0.2058,
|
| 1725 |
+
"step": 12250
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 2.155247941124934,
|
| 1729 |
+
"grad_norm": 0.9375,
|
| 1730 |
+
"learning_rate": 1.4082121371415222e-05,
|
| 1731 |
+
"loss": 0.2087,
|
| 1732 |
+
"step": 12300
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 2.164009111617312,
|
| 1736 |
+
"grad_norm": 1.125,
|
| 1737 |
+
"learning_rate": 1.3936101863208925e-05,
|
| 1738 |
+
"loss": 0.2049,
|
| 1739 |
+
"step": 12350
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 2.1727702821096897,
|
| 1743 |
+
"grad_norm": 1.15625,
|
| 1744 |
+
"learning_rate": 1.3790082355002629e-05,
|
| 1745 |
+
"loss": 0.2073,
|
| 1746 |
+
"step": 12400
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 2.1815314526020675,
|
| 1750 |
+
"grad_norm": 1.203125,
|
| 1751 |
+
"learning_rate": 1.3644062846796332e-05,
|
| 1752 |
+
"loss": 0.2048,
|
| 1753 |
+
"step": 12450
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 2.1902926230944453,
|
| 1757 |
+
"grad_norm": 1.1171875,
|
| 1758 |
+
"learning_rate": 1.3498043338590036e-05,
|
| 1759 |
+
"loss": 0.2062,
|
| 1760 |
+
"step": 12500
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 2.199053793586823,
|
| 1764 |
+
"grad_norm": 1.15625,
|
| 1765 |
+
"learning_rate": 1.3352023830383739e-05,
|
| 1766 |
+
"loss": 0.2104,
|
| 1767 |
+
"step": 12550
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 2.207814964079201,
|
| 1771 |
+
"grad_norm": 1.109375,
|
| 1772 |
+
"learning_rate": 1.3206004322177443e-05,
|
| 1773 |
+
"loss": 0.2075,
|
| 1774 |
+
"step": 12600
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 2.2165761345715786,
|
| 1778 |
+
"grad_norm": 0.9609375,
|
| 1779 |
+
"learning_rate": 1.3059984813971145e-05,
|
| 1780 |
+
"loss": 0.2071,
|
| 1781 |
+
"step": 12650
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 2.2253373050639564,
|
| 1785 |
+
"grad_norm": 1.03125,
|
| 1786 |
+
"learning_rate": 1.291396530576485e-05,
|
| 1787 |
+
"loss": 0.2063,
|
| 1788 |
+
"step": 12700
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 2.2340984755563342,
|
| 1792 |
+
"grad_norm": 1.0625,
|
| 1793 |
+
"learning_rate": 1.2767945797558552e-05,
|
| 1794 |
+
"loss": 0.2033,
|
| 1795 |
+
"step": 12750
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 2.242859646048712,
|
| 1799 |
+
"grad_norm": 1.0625,
|
| 1800 |
+
"learning_rate": 1.262192628935226e-05,
|
| 1801 |
+
"loss": 0.2057,
|
| 1802 |
+
"step": 12800
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 2.25162081654109,
|
| 1806 |
+
"grad_norm": 1.125,
|
| 1807 |
+
"learning_rate": 1.2475906781145961e-05,
|
| 1808 |
+
"loss": 0.2063,
|
| 1809 |
+
"step": 12850
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 2.2603819870334676,
|
| 1813 |
+
"grad_norm": 1.046875,
|
| 1814 |
+
"learning_rate": 1.2329887272939665e-05,
|
| 1815 |
+
"loss": 0.2011,
|
| 1816 |
+
"step": 12900
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 2.2691431575258454,
|
| 1820 |
+
"grad_norm": 1.1328125,
|
| 1821 |
+
"learning_rate": 1.2183867764733368e-05,
|
| 1822 |
+
"loss": 0.2086,
|
| 1823 |
+
"step": 12950
|
| 1824 |
+
},
|
| 1825 |
+
{
|
| 1826 |
+
"epoch": 2.277904328018223,
|
| 1827 |
+
"grad_norm": 1.171875,
|
| 1828 |
+
"learning_rate": 1.2037848256527072e-05,
|
| 1829 |
+
"loss": 0.2063,
|
| 1830 |
+
"step": 13000
|
| 1831 |
+
},
|
| 1832 |
+
{
|
| 1833 |
+
"epoch": 2.286665498510601,
|
| 1834 |
+
"grad_norm": 1.0625,
|
| 1835 |
+
"learning_rate": 1.1891828748320776e-05,
|
| 1836 |
+
"loss": 0.2082,
|
| 1837 |
+
"step": 13050
|
| 1838 |
+
},
|
| 1839 |
+
{
|
| 1840 |
+
"epoch": 2.2954266690029788,
|
| 1841 |
+
"grad_norm": 0.984375,
|
| 1842 |
+
"learning_rate": 1.174580924011448e-05,
|
| 1843 |
+
"loss": 0.2124,
|
| 1844 |
+
"step": 13100
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 2.3041878394953565,
|
| 1848 |
+
"grad_norm": 1.0078125,
|
| 1849 |
+
"learning_rate": 1.1599789731908183e-05,
|
| 1850 |
+
"loss": 0.2078,
|
| 1851 |
+
"step": 13150
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 2.3129490099877343,
|
| 1855 |
+
"grad_norm": 1.046875,
|
| 1856 |
+
"learning_rate": 1.1453770223701888e-05,
|
| 1857 |
+
"loss": 0.2098,
|
| 1858 |
+
"step": 13200
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 2.321710180480112,
|
| 1862 |
+
"grad_norm": 1.046875,
|
| 1863 |
+
"learning_rate": 1.130775071549559e-05,
|
| 1864 |
+
"loss": 0.2057,
|
| 1865 |
+
"step": 13250
|
| 1866 |
+
},
|
| 1867 |
+
{
|
| 1868 |
+
"epoch": 2.33047135097249,
|
| 1869 |
+
"grad_norm": 1.1875,
|
| 1870 |
+
"learning_rate": 1.1161731207289294e-05,
|
| 1871 |
+
"loss": 0.2048,
|
| 1872 |
+
"step": 13300
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 2.3392325214648677,
|
| 1876 |
+
"grad_norm": 1.3203125,
|
| 1877 |
+
"learning_rate": 1.1015711699082999e-05,
|
| 1878 |
+
"loss": 0.2049,
|
| 1879 |
+
"step": 13350
|
| 1880 |
+
},
|
| 1881 |
+
{
|
| 1882 |
+
"epoch": 2.3479936919572455,
|
| 1883 |
+
"grad_norm": 1.0078125,
|
| 1884 |
+
"learning_rate": 1.0869692190876701e-05,
|
| 1885 |
+
"loss": 0.206,
|
| 1886 |
+
"step": 13400
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 2.3567548624496233,
|
| 1890 |
+
"grad_norm": 1.125,
|
| 1891 |
+
"learning_rate": 1.0723672682670406e-05,
|
| 1892 |
+
"loss": 0.2044,
|
| 1893 |
+
"step": 13450
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 2.365516032942001,
|
| 1897 |
+
"grad_norm": 1.2890625,
|
| 1898 |
+
"learning_rate": 1.0577653174464108e-05,
|
| 1899 |
+
"loss": 0.2062,
|
| 1900 |
+
"step": 13500
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"epoch": 2.374277203434379,
|
| 1904 |
+
"grad_norm": 1.140625,
|
| 1905 |
+
"learning_rate": 1.0431633666257812e-05,
|
| 1906 |
+
"loss": 0.206,
|
| 1907 |
+
"step": 13550
|
| 1908 |
+
},
|
| 1909 |
+
{
|
| 1910 |
+
"epoch": 2.3830383739267567,
|
| 1911 |
+
"grad_norm": 1.09375,
|
| 1912 |
+
"learning_rate": 1.0285614158051517e-05,
|
| 1913 |
+
"loss": 0.2044,
|
| 1914 |
+
"step": 13600
|
| 1915 |
+
},
|
| 1916 |
+
{
|
| 1917 |
+
"epoch": 2.3917995444191344,
|
| 1918 |
+
"grad_norm": 1.0546875,
|
| 1919 |
+
"learning_rate": 1.0139594649845221e-05,
|
| 1920 |
+
"loss": 0.207,
|
| 1921 |
+
"step": 13650
|
| 1922 |
+
},
|
| 1923 |
+
{
|
| 1924 |
+
"epoch": 2.4005607149115122,
|
| 1925 |
+
"grad_norm": 1.1953125,
|
| 1926 |
+
"learning_rate": 9.993575141638924e-06,
|
| 1927 |
+
"loss": 0.2032,
|
| 1928 |
+
"step": 13700
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 2.40932188540389,
|
| 1932 |
+
"grad_norm": 1.03125,
|
| 1933 |
+
"learning_rate": 9.847555633432628e-06,
|
| 1934 |
+
"loss": 0.2037,
|
| 1935 |
+
"step": 13750
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 2.418083055896268,
|
| 1939 |
+
"grad_norm": 1.2265625,
|
| 1940 |
+
"learning_rate": 9.70153612522633e-06,
|
| 1941 |
+
"loss": 0.2076,
|
| 1942 |
+
"step": 13800
|
| 1943 |
+
},
|
| 1944 |
+
{
|
| 1945 |
+
"epoch": 2.4268442263886456,
|
| 1946 |
+
"grad_norm": 0.984375,
|
| 1947 |
+
"learning_rate": 9.555516617020035e-06,
|
| 1948 |
+
"loss": 0.2049,
|
| 1949 |
+
"step": 13850
|
| 1950 |
+
},
|
| 1951 |
+
{
|
| 1952 |
+
"epoch": 2.4356053968810234,
|
| 1953 |
+
"grad_norm": 1.125,
|
| 1954 |
+
"learning_rate": 9.409497108813737e-06,
|
| 1955 |
+
"loss": 0.2046,
|
| 1956 |
+
"step": 13900
|
| 1957 |
+
},
|
| 1958 |
+
{
|
| 1959 |
+
"epoch": 2.444366567373401,
|
| 1960 |
+
"grad_norm": 1.171875,
|
| 1961 |
+
"learning_rate": 9.263477600607442e-06,
|
| 1962 |
+
"loss": 0.2044,
|
| 1963 |
+
"step": 13950
|
| 1964 |
+
},
|
| 1965 |
+
{
|
| 1966 |
+
"epoch": 2.453127737865779,
|
| 1967 |
+
"grad_norm": 1.15625,
|
| 1968 |
+
"learning_rate": 9.117458092401144e-06,
|
| 1969 |
+
"loss": 0.2061,
|
| 1970 |
+
"step": 14000
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 2.4618889083581568,
|
| 1974 |
+
"grad_norm": 1.1328125,
|
| 1975 |
+
"learning_rate": 8.97143858419485e-06,
|
| 1976 |
+
"loss": 0.206,
|
| 1977 |
+
"step": 14050
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 2.4706500788505346,
|
| 1981 |
+
"grad_norm": 1.0390625,
|
| 1982 |
+
"learning_rate": 8.825419075988553e-06,
|
| 1983 |
+
"loss": 0.2012,
|
| 1984 |
+
"step": 14100
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 2.4794112493429123,
|
| 1988 |
+
"grad_norm": 1.0,
|
| 1989 |
+
"learning_rate": 8.679399567782257e-06,
|
| 1990 |
+
"loss": 0.2008,
|
| 1991 |
+
"step": 14150
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"epoch": 2.48817241983529,
|
| 1995 |
+
"grad_norm": 1.1328125,
|
| 1996 |
+
"learning_rate": 8.53338005957596e-06,
|
| 1997 |
+
"loss": 0.2046,
|
| 1998 |
+
"step": 14200
|
| 1999 |
+
},
|
| 2000 |
+
{
|
| 2001 |
+
"epoch": 2.496933590327668,
|
| 2002 |
+
"grad_norm": 1.1328125,
|
| 2003 |
+
"learning_rate": 8.387360551369664e-06,
|
| 2004 |
+
"loss": 0.2067,
|
| 2005 |
+
"step": 14250
|
| 2006 |
+
},
|
| 2007 |
+
{
|
| 2008 |
+
"epoch": 2.5056947608200457,
|
| 2009 |
+
"grad_norm": 1.2109375,
|
| 2010 |
+
"learning_rate": 8.241341043163366e-06,
|
| 2011 |
+
"loss": 0.2041,
|
| 2012 |
+
"step": 14300
|
| 2013 |
+
},
|
| 2014 |
+
{
|
| 2015 |
+
"epoch": 2.5144559313124235,
|
| 2016 |
+
"grad_norm": 1.171875,
|
| 2017 |
+
"learning_rate": 8.09532153495707e-06,
|
| 2018 |
+
"loss": 0.2066,
|
| 2019 |
+
"step": 14350
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 2.523217101804801,
|
| 2023 |
+
"grad_norm": 0.984375,
|
| 2024 |
+
"learning_rate": 7.949302026750773e-06,
|
| 2025 |
+
"loss": 0.2036,
|
| 2026 |
+
"step": 14400
|
| 2027 |
+
},
|
| 2028 |
+
{
|
| 2029 |
+
"epoch": 2.531978272297179,
|
| 2030 |
+
"grad_norm": 1.1640625,
|
| 2031 |
+
"learning_rate": 7.803282518544478e-06,
|
| 2032 |
+
"loss": 0.2002,
|
| 2033 |
+
"step": 14450
|
| 2034 |
+
},
|
| 2035 |
+
{
|
| 2036 |
+
"epoch": 2.5407394427895564,
|
| 2037 |
+
"grad_norm": 1.203125,
|
| 2038 |
+
"learning_rate": 7.65726301033818e-06,
|
| 2039 |
+
"loss": 0.2057,
|
| 2040 |
+
"step": 14500
|
| 2041 |
+
},
|
| 2042 |
+
{
|
| 2043 |
+
"epoch": 2.5495006132819347,
|
| 2044 |
+
"grad_norm": 1.1015625,
|
| 2045 |
+
"learning_rate": 7.511243502131886e-06,
|
| 2046 |
+
"loss": 0.2069,
|
| 2047 |
+
"step": 14550
|
| 2048 |
+
},
|
| 2049 |
+
{
|
| 2050 |
+
"epoch": 2.558261783774312,
|
| 2051 |
+
"grad_norm": 1.28125,
|
| 2052 |
+
"learning_rate": 7.36522399392559e-06,
|
| 2053 |
+
"loss": 0.1996,
|
| 2054 |
+
"step": 14600
|
| 2055 |
+
},
|
| 2056 |
+
{
|
| 2057 |
+
"epoch": 2.5670229542666902,
|
| 2058 |
+
"grad_norm": 1.09375,
|
| 2059 |
+
"learning_rate": 7.219204485719293e-06,
|
| 2060 |
+
"loss": 0.2085,
|
| 2061 |
+
"step": 14650
|
| 2062 |
+
},
|
| 2063 |
+
{
|
| 2064 |
+
"epoch": 2.5757841247590676,
|
| 2065 |
+
"grad_norm": 1.0703125,
|
| 2066 |
+
"learning_rate": 7.0731849775129965e-06,
|
| 2067 |
+
"loss": 0.2068,
|
| 2068 |
+
"step": 14700
|
| 2069 |
+
},
|
| 2070 |
+
{
|
| 2071 |
+
"epoch": 2.584545295251446,
|
| 2072 |
+
"grad_norm": 1.046875,
|
| 2073 |
+
"learning_rate": 6.9271654693067e-06,
|
| 2074 |
+
"loss": 0.2054,
|
| 2075 |
+
"step": 14750
|
| 2076 |
+
},
|
| 2077 |
+
{
|
| 2078 |
+
"epoch": 2.593306465743823,
|
| 2079 |
+
"grad_norm": 1.1171875,
|
| 2080 |
+
"learning_rate": 6.781145961100403e-06,
|
| 2081 |
+
"loss": 0.2051,
|
| 2082 |
+
"step": 14800
|
| 2083 |
+
},
|
| 2084 |
+
{
|
| 2085 |
+
"epoch": 2.6020676362362014,
|
| 2086 |
+
"grad_norm": 1.15625,
|
| 2087 |
+
"learning_rate": 6.635126452894107e-06,
|
| 2088 |
+
"loss": 0.2063,
|
| 2089 |
+
"step": 14850
|
| 2090 |
+
},
|
| 2091 |
+
{
|
| 2092 |
+
"epoch": 2.6108288067285788,
|
| 2093 |
+
"grad_norm": 1.0859375,
|
| 2094 |
+
"learning_rate": 6.48910694468781e-06,
|
| 2095 |
+
"loss": 0.208,
|
| 2096 |
+
"step": 14900
|
| 2097 |
+
},
|
| 2098 |
+
{
|
| 2099 |
+
"epoch": 2.619589977220957,
|
| 2100 |
+
"grad_norm": 1.21875,
|
| 2101 |
+
"learning_rate": 6.343087436481514e-06,
|
| 2102 |
+
"loss": 0.2069,
|
| 2103 |
+
"step": 14950
|
| 2104 |
+
},
|
| 2105 |
+
{
|
| 2106 |
+
"epoch": 2.6283511477133343,
|
| 2107 |
+
"grad_norm": 1.09375,
|
| 2108 |
+
"learning_rate": 6.197067928275218e-06,
|
| 2109 |
+
"loss": 0.2033,
|
| 2110 |
+
"step": 15000
|
| 2111 |
+
},
|
| 2112 |
+
{
|
| 2113 |
+
"epoch": 2.637112318205712,
|
| 2114 |
+
"grad_norm": 1.109375,
|
| 2115 |
+
"learning_rate": 6.0510484200689214e-06,
|
| 2116 |
+
"loss": 0.2058,
|
| 2117 |
+
"step": 15050
|
| 2118 |
+
},
|
| 2119 |
+
{
|
| 2120 |
+
"epoch": 2.64587348869809,
|
| 2121 |
+
"grad_norm": 1.0859375,
|
| 2122 |
+
"learning_rate": 5.905028911862625e-06,
|
| 2123 |
+
"loss": 0.2044,
|
| 2124 |
+
"step": 15100
|
| 2125 |
+
},
|
| 2126 |
+
{
|
| 2127 |
+
"epoch": 2.6546346591904677,
|
| 2128 |
+
"grad_norm": 1.09375,
|
| 2129 |
+
"learning_rate": 5.759009403656328e-06,
|
| 2130 |
+
"loss": 0.2018,
|
| 2131 |
+
"step": 15150
|
| 2132 |
+
},
|
| 2133 |
+
{
|
| 2134 |
+
"epoch": 2.6633958296828455,
|
| 2135 |
+
"grad_norm": 1.078125,
|
| 2136 |
+
"learning_rate": 5.612989895450033e-06,
|
| 2137 |
+
"loss": 0.2032,
|
| 2138 |
+
"step": 15200
|
| 2139 |
+
},
|
| 2140 |
+
{
|
| 2141 |
+
"epoch": 2.6721570001752233,
|
| 2142 |
+
"grad_norm": 1.015625,
|
| 2143 |
+
"learning_rate": 5.466970387243736e-06,
|
| 2144 |
+
"loss": 0.206,
|
| 2145 |
+
"step": 15250
|
| 2146 |
+
},
|
| 2147 |
+
{
|
| 2148 |
+
"epoch": 2.680918170667601,
|
| 2149 |
+
"grad_norm": 1.09375,
|
| 2150 |
+
"learning_rate": 5.3209508790374395e-06,
|
| 2151 |
+
"loss": 0.2051,
|
| 2152 |
+
"step": 15300
|
| 2153 |
+
},
|
| 2154 |
+
{
|
| 2155 |
+
"epoch": 2.689679341159979,
|
| 2156 |
+
"grad_norm": 1.0546875,
|
| 2157 |
+
"learning_rate": 5.174931370831143e-06,
|
| 2158 |
+
"loss": 0.2048,
|
| 2159 |
+
"step": 15350
|
| 2160 |
+
},
|
| 2161 |
+
{
|
| 2162 |
+
"epoch": 2.6984405116523567,
|
| 2163 |
+
"grad_norm": 1.2421875,
|
| 2164 |
+
"learning_rate": 5.028911862624846e-06,
|
| 2165 |
+
"loss": 0.2074,
|
| 2166 |
+
"step": 15400
|
| 2167 |
+
},
|
| 2168 |
+
{
|
| 2169 |
+
"epoch": 2.7072016821447344,
|
| 2170 |
+
"grad_norm": 1.0390625,
|
| 2171 |
+
"learning_rate": 4.882892354418551e-06,
|
| 2172 |
+
"loss": 0.2043,
|
| 2173 |
+
"step": 15450
|
| 2174 |
+
},
|
| 2175 |
+
{
|
| 2176 |
+
"epoch": 2.7159628526371122,
|
| 2177 |
+
"grad_norm": 1.0703125,
|
| 2178 |
+
"learning_rate": 4.736872846212254e-06,
|
| 2179 |
+
"loss": 0.2064,
|
| 2180 |
+
"step": 15500
|
| 2181 |
+
},
|
| 2182 |
+
{
|
| 2183 |
+
"epoch": 2.72472402312949,
|
| 2184 |
+
"grad_norm": 1.140625,
|
| 2185 |
+
"learning_rate": 4.5908533380059575e-06,
|
| 2186 |
+
"loss": 0.2044,
|
| 2187 |
+
"step": 15550
|
| 2188 |
+
},
|
| 2189 |
+
{
|
| 2190 |
+
"epoch": 2.733485193621868,
|
| 2191 |
+
"grad_norm": 1.0390625,
|
| 2192 |
+
"learning_rate": 4.444833829799661e-06,
|
| 2193 |
+
"loss": 0.2078,
|
| 2194 |
+
"step": 15600
|
| 2195 |
+
},
|
| 2196 |
+
{
|
| 2197 |
+
"epoch": 2.7422463641142456,
|
| 2198 |
+
"grad_norm": 1.2265625,
|
| 2199 |
+
"learning_rate": 4.298814321593364e-06,
|
| 2200 |
+
"loss": 0.2018,
|
| 2201 |
+
"step": 15650
|
| 2202 |
+
},
|
| 2203 |
+
{
|
| 2204 |
+
"epoch": 2.7510075346066234,
|
| 2205 |
+
"grad_norm": 1.15625,
|
| 2206 |
+
"learning_rate": 4.152794813387069e-06,
|
| 2207 |
+
"loss": 0.2072,
|
| 2208 |
+
"step": 15700
|
| 2209 |
+
},
|
| 2210 |
+
{
|
| 2211 |
+
"epoch": 2.759768705099001,
|
| 2212 |
+
"grad_norm": 1.1484375,
|
| 2213 |
+
"learning_rate": 4.006775305180772e-06,
|
| 2214 |
+
"loss": 0.2054,
|
| 2215 |
+
"step": 15750
|
| 2216 |
+
},
|
| 2217 |
+
{
|
| 2218 |
+
"epoch": 2.768529875591379,
|
| 2219 |
+
"grad_norm": 1.1328125,
|
| 2220 |
+
"learning_rate": 3.8607557969744755e-06,
|
| 2221 |
+
"loss": 0.2067,
|
| 2222 |
+
"step": 15800
|
| 2223 |
+
},
|
| 2224 |
+
{
|
| 2225 |
+
"epoch": 2.7772910460837568,
|
| 2226 |
+
"grad_norm": 1.234375,
|
| 2227 |
+
"learning_rate": 3.7147362887681794e-06,
|
| 2228 |
+
"loss": 0.2029,
|
| 2229 |
+
"step": 15850
|
| 2230 |
+
},
|
| 2231 |
+
{
|
| 2232 |
+
"epoch": 2.7860522165761346,
|
| 2233 |
+
"grad_norm": 1.125,
|
| 2234 |
+
"learning_rate": 3.568716780561883e-06,
|
| 2235 |
+
"loss": 0.2006,
|
| 2236 |
+
"step": 15900
|
| 2237 |
+
},
|
| 2238 |
+
{
|
| 2239 |
+
"epoch": 2.7948133870685123,
|
| 2240 |
+
"grad_norm": 1.171875,
|
| 2241 |
+
"learning_rate": 3.422697272355587e-06,
|
| 2242 |
+
"loss": 0.2032,
|
| 2243 |
+
"step": 15950
|
| 2244 |
+
},
|
| 2245 |
+
{
|
| 2246 |
+
"epoch": 2.80357455756089,
|
| 2247 |
+
"grad_norm": 1.015625,
|
| 2248 |
+
"learning_rate": 3.2766777641492905e-06,
|
| 2249 |
+
"loss": 0.2056,
|
| 2250 |
+
"step": 16000
|
| 2251 |
+
},
|
| 2252 |
+
{
|
| 2253 |
+
"epoch": 2.812335728053268,
|
| 2254 |
+
"grad_norm": 1.1484375,
|
| 2255 |
+
"learning_rate": 3.130658255942994e-06,
|
| 2256 |
+
"loss": 0.2069,
|
| 2257 |
+
"step": 16050
|
| 2258 |
+
},
|
| 2259 |
+
{
|
| 2260 |
+
"epoch": 2.8210968985456457,
|
| 2261 |
+
"grad_norm": 1.1171875,
|
| 2262 |
+
"learning_rate": 2.984638747736698e-06,
|
| 2263 |
+
"loss": 0.2057,
|
| 2264 |
+
"step": 16100
|
| 2265 |
+
},
|
| 2266 |
+
{
|
| 2267 |
+
"epoch": 2.8298580690380235,
|
| 2268 |
+
"grad_norm": 1.125,
|
| 2269 |
+
"learning_rate": 2.8386192395304013e-06,
|
| 2270 |
+
"loss": 0.2069,
|
| 2271 |
+
"step": 16150
|
| 2272 |
+
},
|
| 2273 |
+
{
|
| 2274 |
+
"epoch": 2.8386192395304013,
|
| 2275 |
+
"grad_norm": 1.109375,
|
| 2276 |
+
"learning_rate": 2.6925997313241047e-06,
|
| 2277 |
+
"loss": 0.2047,
|
| 2278 |
+
"step": 16200
|
| 2279 |
+
},
|
| 2280 |
+
{
|
| 2281 |
+
"epoch": 2.847380410022779,
|
| 2282 |
+
"grad_norm": 1.0625,
|
| 2283 |
+
"learning_rate": 2.5465802231178086e-06,
|
| 2284 |
+
"loss": 0.2093,
|
| 2285 |
+
"step": 16250
|
| 2286 |
+
},
|
| 2287 |
+
{
|
| 2288 |
+
"epoch": 2.856141580515157,
|
| 2289 |
+
"grad_norm": 1.015625,
|
| 2290 |
+
"learning_rate": 2.4005607149115124e-06,
|
| 2291 |
+
"loss": 0.2047,
|
| 2292 |
+
"step": 16300
|
| 2293 |
+
},
|
| 2294 |
+
{
|
| 2295 |
+
"epoch": 2.8649027510075347,
|
| 2296 |
+
"grad_norm": 1.1953125,
|
| 2297 |
+
"learning_rate": 2.2545412067052163e-06,
|
| 2298 |
+
"loss": 0.2042,
|
| 2299 |
+
"step": 16350
|
| 2300 |
+
},
|
| 2301 |
+
{
|
| 2302 |
+
"epoch": 2.8736639214999125,
|
| 2303 |
+
"grad_norm": 1.1953125,
|
| 2304 |
+
"learning_rate": 2.1085216984989197e-06,
|
| 2305 |
+
"loss": 0.2051,
|
| 2306 |
+
"step": 16400
|
| 2307 |
+
},
|
| 2308 |
+
{
|
| 2309 |
+
"epoch": 2.8824250919922902,
|
| 2310 |
+
"grad_norm": 1.2734375,
|
| 2311 |
+
"learning_rate": 1.962502190292623e-06,
|
| 2312 |
+
"loss": 0.2026,
|
| 2313 |
+
"step": 16450
|
| 2314 |
+
},
|
| 2315 |
+
{
|
| 2316 |
+
"epoch": 2.891186262484668,
|
| 2317 |
+
"grad_norm": 1.203125,
|
| 2318 |
+
"learning_rate": 1.8164826820863268e-06,
|
| 2319 |
+
"loss": 0.2134,
|
| 2320 |
+
"step": 16500
|
| 2321 |
+
},
|
| 2322 |
+
{
|
| 2323 |
+
"epoch": 2.899947432977046,
|
| 2324 |
+
"grad_norm": 1.265625,
|
| 2325 |
+
"learning_rate": 1.6704631738800305e-06,
|
| 2326 |
+
"loss": 0.2035,
|
| 2327 |
+
"step": 16550
|
| 2328 |
+
},
|
| 2329 |
+
{
|
| 2330 |
+
"epoch": 2.9087086034694236,
|
| 2331 |
+
"grad_norm": 1.0078125,
|
| 2332 |
+
"learning_rate": 1.524443665673734e-06,
|
| 2333 |
+
"loss": 0.2024,
|
| 2334 |
+
"step": 16600
|
| 2335 |
+
},
|
| 2336 |
+
{
|
| 2337 |
+
"epoch": 2.9174697739618014,
|
| 2338 |
+
"grad_norm": 1.1484375,
|
| 2339 |
+
"learning_rate": 1.3784241574674377e-06,
|
| 2340 |
+
"loss": 0.209,
|
| 2341 |
+
"step": 16650
|
| 2342 |
+
},
|
| 2343 |
+
{
|
| 2344 |
+
"epoch": 2.926230944454179,
|
| 2345 |
+
"grad_norm": 1.0234375,
|
| 2346 |
+
"learning_rate": 1.2324046492611414e-06,
|
| 2347 |
+
"loss": 0.2035,
|
| 2348 |
+
"step": 16700
|
| 2349 |
+
},
|
| 2350 |
+
{
|
| 2351 |
+
"epoch": 2.934992114946557,
|
| 2352 |
+
"grad_norm": 1.15625,
|
| 2353 |
+
"learning_rate": 1.0863851410548448e-06,
|
| 2354 |
+
"loss": 0.2061,
|
| 2355 |
+
"step": 16750
|
| 2356 |
+
},
|
| 2357 |
+
{
|
| 2358 |
+
"epoch": 2.9437532854389348,
|
| 2359 |
+
"grad_norm": 1.140625,
|
| 2360 |
+
"learning_rate": 9.403656328485486e-07,
|
| 2361 |
+
"loss": 0.2049,
|
| 2362 |
+
"step": 16800
|
| 2363 |
+
},
|
| 2364 |
+
{
|
| 2365 |
+
"epoch": 2.9525144559313126,
|
| 2366 |
+
"grad_norm": 0.98046875,
|
| 2367 |
+
"learning_rate": 7.943461246422522e-07,
|
| 2368 |
+
"loss": 0.2058,
|
| 2369 |
+
"step": 16850
|
| 2370 |
+
},
|
| 2371 |
+
{
|
| 2372 |
+
"epoch": 2.9612756264236904,
|
| 2373 |
+
"grad_norm": 1.046875,
|
| 2374 |
+
"learning_rate": 6.483266164359559e-07,
|
| 2375 |
+
"loss": 0.2049,
|
| 2376 |
+
"step": 16900
|
| 2377 |
+
},
|
| 2378 |
+
{
|
| 2379 |
+
"epoch": 2.970036796916068,
|
| 2380 |
+
"grad_norm": 1.109375,
|
| 2381 |
+
"learning_rate": 5.023071082296594e-07,
|
| 2382 |
+
"loss": 0.2014,
|
| 2383 |
+
"step": 16950
|
| 2384 |
+
},
|
| 2385 |
+
{
|
| 2386 |
+
"epoch": 2.978797967408446,
|
| 2387 |
+
"grad_norm": 1.0546875,
|
| 2388 |
+
"learning_rate": 3.562876000233631e-07,
|
| 2389 |
+
"loss": 0.2054,
|
| 2390 |
+
"step": 17000
|
| 2391 |
+
},
|
| 2392 |
+
{
|
| 2393 |
+
"epoch": 2.9875591379008233,
|
| 2394 |
+
"grad_norm": 1.015625,
|
| 2395 |
+
"learning_rate": 2.1026809181706677e-07,
|
| 2396 |
+
"loss": 0.2021,
|
| 2397 |
+
"step": 17050
|
| 2398 |
+
},
|
| 2399 |
+
{
|
| 2400 |
+
"epoch": 2.9963203083932015,
|
| 2401 |
+
"grad_norm": 1.03125,
|
| 2402 |
+
"learning_rate": 6.42485836107704e-08,
|
| 2403 |
+
"loss": 0.2041,
|
| 2404 |
+
"step": 17100
|
| 2405 |
+
}
|
| 2406 |
+
],
|
| 2407 |
+
"logging_steps": 50,
|
| 2408 |
+
"max_steps": 17121,
|
| 2409 |
+
"num_input_tokens_seen": 0,
|
| 2410 |
+
"num_train_epochs": 3,
|
| 2411 |
+
"save_steps": 500,
|
| 2412 |
+
"stateful_callbacks": {
|
| 2413 |
+
"TrainerControl": {
|
| 2414 |
+
"args": {
|
| 2415 |
+
"should_epoch_stop": false,
|
| 2416 |
+
"should_evaluate": false,
|
| 2417 |
+
"should_log": false,
|
| 2418 |
+
"should_save": true,
|
| 2419 |
+
"should_training_stop": true
|
| 2420 |
+
},
|
| 2421 |
+
"attributes": {}
|
| 2422 |
+
}
|
| 2423 |
+
},
|
| 2424 |
+
"total_flos": 1.174577165177979e+18,
|
| 2425 |
+
"train_batch_size": 8,
|
| 2426 |
+
"trial_name": null,
|
| 2427 |
+
"trial_params": null
|
| 2428 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fb53905ef88322abfa3f7e9d62ab12a7ead39da42e53470b6a2c770651026fe1
|
| 3 |
+
size 5304
|