Upload folder using huggingface_hub
Browse files- config.json +35 -0
- generation_config.json +9 -0
- model.safetensors +3 -0
- optimizer.pt +3 -0
- rng_state.pth +3 -0
- scaler.pt +3 -0
- scheduler.pt +3 -0
- trainer_state.json +1217 -0
- training_args.bin +3 -0
config.json
ADDED
|
@@ -0,0 +1,35 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"activation_function": "gelu_new",
|
| 3 |
+
"add_cross_attention": false,
|
| 4 |
+
"architectures": [
|
| 5 |
+
"GPT2LMHeadModel"
|
| 6 |
+
],
|
| 7 |
+
"attn_pdrop": 0.1,
|
| 8 |
+
"bos_token_id": 50256,
|
| 9 |
+
"dtype": "float32",
|
| 10 |
+
"embd_pdrop": 0.1,
|
| 11 |
+
"eos_token_id": 50256,
|
| 12 |
+
"initializer_range": 0.02,
|
| 13 |
+
"layer_norm_epsilon": 1e-05,
|
| 14 |
+
"model_type": "gpt2",
|
| 15 |
+
"n_ctx": 1024,
|
| 16 |
+
"n_embd": 512,
|
| 17 |
+
"n_head": 4,
|
| 18 |
+
"n_inner": null,
|
| 19 |
+
"n_layer": 4,
|
| 20 |
+
"n_positions": 1024,
|
| 21 |
+
"pad_token_id": null,
|
| 22 |
+
"reorder_and_upcast_attn": false,
|
| 23 |
+
"resid_pdrop": 0.1,
|
| 24 |
+
"scale_attn_by_inverse_layer_idx": false,
|
| 25 |
+
"scale_attn_weights": true,
|
| 26 |
+
"summary_activation": null,
|
| 27 |
+
"summary_first_dropout": 0.1,
|
| 28 |
+
"summary_proj_to_labels": true,
|
| 29 |
+
"summary_type": "cls_index",
|
| 30 |
+
"summary_use_proj": true,
|
| 31 |
+
"tie_word_embeddings": true,
|
| 32 |
+
"transformers_version": "5.14.1",
|
| 33 |
+
"use_cache": false,
|
| 34 |
+
"vocab_size": 50257
|
| 35 |
+
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 50256,
|
| 4 |
+
"eos_token_id": 50256,
|
| 5 |
+
"output_attentions": false,
|
| 6 |
+
"output_hidden_states": false,
|
| 7 |
+
"transformers_version": "5.14.1",
|
| 8 |
+
"use_cache": true
|
| 9 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b73e227fa51d2bcd33d10c359652e1ec831d0e1da81da86d9fa2b1a74b9c0f6a
|
| 3 |
+
size 155470896
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0571c060ac40e58db39d06c0875855040beb9e2ee7cd35cbcf667dc274319292
|
| 3 |
+
size 310976459
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0e73a2f3a33fc6f2835cd718fdd27fda404df6a31df34ee8e303987b08c218da
|
| 3 |
+
size 14645
|
scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e255737f3f46a218d397a90126866638ed501fbb8586ea4a228d63947f407a11
|
| 3 |
+
size 1383
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c5678c27b6983c3834d758b23695ed33c1daaeefccf77db24d8e350bdfc8fbf9
|
| 3 |
+
size 1465
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1217 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 10.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1690,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.05917159763313609,
|
| 14 |
+
"grad_norm": 2.13480806350708,
|
| 15 |
+
"learning_rate": 0.0004973372781065089,
|
| 16 |
+
"loss": 8.609089660644532,
|
| 17 |
+
"step": 10
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.11834319526627218,
|
| 21 |
+
"grad_norm": 0.5722453594207764,
|
| 22 |
+
"learning_rate": 0.0004943786982248521,
|
| 23 |
+
"loss": 6.307852554321289,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.17751479289940827,
|
| 28 |
+
"grad_norm": 0.5311302542686462,
|
| 29 |
+
"learning_rate": 0.0004914201183431953,
|
| 30 |
+
"loss": 5.795976257324218,
|
| 31 |
+
"step": 30
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.23668639053254437,
|
| 35 |
+
"grad_norm": 0.936571478843689,
|
| 36 |
+
"learning_rate": 0.0004884615384615384,
|
| 37 |
+
"loss": 5.6642906188964846,
|
| 38 |
+
"step": 40
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.2958579881656805,
|
| 42 |
+
"grad_norm": 0.46537888050079346,
|
| 43 |
+
"learning_rate": 0.0004855029585798817,
|
| 44 |
+
"loss": 5.564136123657226,
|
| 45 |
+
"step": 50
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.35502958579881655,
|
| 49 |
+
"grad_norm": 0.5137708783149719,
|
| 50 |
+
"learning_rate": 0.00048254437869822486,
|
| 51 |
+
"loss": 5.482704925537109,
|
| 52 |
+
"step": 60
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.41420118343195267,
|
| 56 |
+
"grad_norm": 0.6558282971382141,
|
| 57 |
+
"learning_rate": 0.00047958579881656805,
|
| 58 |
+
"loss": 5.424795532226563,
|
| 59 |
+
"step": 70
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.47337278106508873,
|
| 63 |
+
"grad_norm": 0.602185845375061,
|
| 64 |
+
"learning_rate": 0.00047662721893491125,
|
| 65 |
+
"loss": 5.312729263305664,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.5325443786982249,
|
| 70 |
+
"grad_norm": 0.49187108874320984,
|
| 71 |
+
"learning_rate": 0.0004736686390532545,
|
| 72 |
+
"loss": 5.2327423095703125,
|
| 73 |
+
"step": 90
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.591715976331361,
|
| 77 |
+
"grad_norm": 0.5671802163124084,
|
| 78 |
+
"learning_rate": 0.00047071005917159763,
|
| 79 |
+
"loss": 5.180327224731445,
|
| 80 |
+
"step": 100
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.650887573964497,
|
| 84 |
+
"grad_norm": 0.5761376619338989,
|
| 85 |
+
"learning_rate": 0.0004677514792899408,
|
| 86 |
+
"loss": 5.1113639831542965,
|
| 87 |
+
"step": 110
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.7100591715976331,
|
| 91 |
+
"grad_norm": 0.6834049820899963,
|
| 92 |
+
"learning_rate": 0.000464792899408284,
|
| 93 |
+
"loss": 5.110577011108399,
|
| 94 |
+
"step": 120
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.7692307692307693,
|
| 98 |
+
"grad_norm": 0.4934782385826111,
|
| 99 |
+
"learning_rate": 0.00046183431952662727,
|
| 100 |
+
"loss": 5.033809661865234,
|
| 101 |
+
"step": 130
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.8284023668639053,
|
| 105 |
+
"grad_norm": 0.5475414395332336,
|
| 106 |
+
"learning_rate": 0.0004588757396449704,
|
| 107 |
+
"loss": 5.023288726806641,
|
| 108 |
+
"step": 140
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.8875739644970414,
|
| 112 |
+
"grad_norm": 0.6682299971580505,
|
| 113 |
+
"learning_rate": 0.0004559171597633136,
|
| 114 |
+
"loss": 5.008454895019531,
|
| 115 |
+
"step": 150
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.9467455621301775,
|
| 119 |
+
"grad_norm": 0.6575542092323303,
|
| 120 |
+
"learning_rate": 0.00045295857988165685,
|
| 121 |
+
"loss": 5.023602676391602,
|
| 122 |
+
"step": 160
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 1.0059171597633136,
|
| 126 |
+
"grad_norm": 0.5498866438865662,
|
| 127 |
+
"learning_rate": 0.00045000000000000004,
|
| 128 |
+
"loss": 4.9478710174560545,
|
| 129 |
+
"step": 170
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 1.0650887573964498,
|
| 133 |
+
"grad_norm": 0.5482797026634216,
|
| 134 |
+
"learning_rate": 0.0004470414201183432,
|
| 135 |
+
"loss": 4.826501083374024,
|
| 136 |
+
"step": 180
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 1.1242603550295858,
|
| 140 |
+
"grad_norm": 0.473796546459198,
|
| 141 |
+
"learning_rate": 0.0004440828402366864,
|
| 142 |
+
"loss": 4.809365081787109,
|
| 143 |
+
"step": 190
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 1.183431952662722,
|
| 147 |
+
"grad_norm": 0.5151321291923523,
|
| 148 |
+
"learning_rate": 0.0004411242603550296,
|
| 149 |
+
"loss": 4.864156723022461,
|
| 150 |
+
"step": 200
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 1.242603550295858,
|
| 154 |
+
"grad_norm": 0.6074804663658142,
|
| 155 |
+
"learning_rate": 0.0004381656804733728,
|
| 156 |
+
"loss": 4.838579559326172,
|
| 157 |
+
"step": 210
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 1.301775147928994,
|
| 161 |
+
"grad_norm": 0.5551325678825378,
|
| 162 |
+
"learning_rate": 0.00043520710059171596,
|
| 163 |
+
"loss": 4.841543579101563,
|
| 164 |
+
"step": 220
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 1.3609467455621302,
|
| 168 |
+
"grad_norm": 0.566065788269043,
|
| 169 |
+
"learning_rate": 0.00043224852071005915,
|
| 170 |
+
"loss": 4.785538864135742,
|
| 171 |
+
"step": 230
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 1.4201183431952662,
|
| 175 |
+
"grad_norm": 0.5819883942604065,
|
| 176 |
+
"learning_rate": 0.0004292899408284024,
|
| 177 |
+
"loss": 4.791210556030274,
|
| 178 |
+
"step": 240
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 1.4792899408284024,
|
| 182 |
+
"grad_norm": 0.6059792637825012,
|
| 183 |
+
"learning_rate": 0.0004263313609467456,
|
| 184 |
+
"loss": 4.756925582885742,
|
| 185 |
+
"step": 250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 1.5384615384615383,
|
| 189 |
+
"grad_norm": 0.7058536410331726,
|
| 190 |
+
"learning_rate": 0.00042337278106508873,
|
| 191 |
+
"loss": 4.7325084686279295,
|
| 192 |
+
"step": 260
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 1.5976331360946747,
|
| 196 |
+
"grad_norm": 0.5290840864181519,
|
| 197 |
+
"learning_rate": 0.000420414201183432,
|
| 198 |
+
"loss": 4.736005020141602,
|
| 199 |
+
"step": 270
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 1.6568047337278107,
|
| 203 |
+
"grad_norm": 0.543929934501648,
|
| 204 |
+
"learning_rate": 0.00041745562130177517,
|
| 205 |
+
"loss": 4.721459197998047,
|
| 206 |
+
"step": 280
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 1.7159763313609466,
|
| 210 |
+
"grad_norm": 0.587887704372406,
|
| 211 |
+
"learning_rate": 0.00041449704142011837,
|
| 212 |
+
"loss": 4.697582244873047,
|
| 213 |
+
"step": 290
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 1.7751479289940828,
|
| 217 |
+
"grad_norm": 0.6189513802528381,
|
| 218 |
+
"learning_rate": 0.0004115384615384615,
|
| 219 |
+
"loss": 4.6849010467529295,
|
| 220 |
+
"step": 300
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 1.834319526627219,
|
| 224 |
+
"grad_norm": 0.6452702879905701,
|
| 225 |
+
"learning_rate": 0.00040857988165680475,
|
| 226 |
+
"loss": 4.6623176574707035,
|
| 227 |
+
"step": 310
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 1.893491124260355,
|
| 231 |
+
"grad_norm": 0.593474805355072,
|
| 232 |
+
"learning_rate": 0.00040562130177514795,
|
| 233 |
+
"loss": 4.680774688720703,
|
| 234 |
+
"step": 320
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 1.952662721893491,
|
| 238 |
+
"grad_norm": 0.5981247425079346,
|
| 239 |
+
"learning_rate": 0.00040266272189349114,
|
| 240 |
+
"loss": 4.683719253540039,
|
| 241 |
+
"step": 330
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 2.0118343195266273,
|
| 245 |
+
"grad_norm": 0.6405901312828064,
|
| 246 |
+
"learning_rate": 0.0003997041420118343,
|
| 247 |
+
"loss": 4.630902481079102,
|
| 248 |
+
"step": 340
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 2.0710059171597632,
|
| 252 |
+
"grad_norm": 0.5495120286941528,
|
| 253 |
+
"learning_rate": 0.0003967455621301775,
|
| 254 |
+
"loss": 4.565330505371094,
|
| 255 |
+
"step": 350
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 2.1301775147928996,
|
| 259 |
+
"grad_norm": 0.5600634813308716,
|
| 260 |
+
"learning_rate": 0.0003937869822485207,
|
| 261 |
+
"loss": 4.529523468017578,
|
| 262 |
+
"step": 360
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 2.1893491124260356,
|
| 266 |
+
"grad_norm": 0.5625339150428772,
|
| 267 |
+
"learning_rate": 0.0003908284023668639,
|
| 268 |
+
"loss": 4.531766128540039,
|
| 269 |
+
"step": 370
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 2.2485207100591715,
|
| 273 |
+
"grad_norm": 0.6504446864128113,
|
| 274 |
+
"learning_rate": 0.00038786982248520716,
|
| 275 |
+
"loss": 4.5285888671875,
|
| 276 |
+
"step": 380
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 2.3076923076923075,
|
| 280 |
+
"grad_norm": 0.5576902627944946,
|
| 281 |
+
"learning_rate": 0.0003849112426035503,
|
| 282 |
+
"loss": 4.525382232666016,
|
| 283 |
+
"step": 390
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 2.366863905325444,
|
| 287 |
+
"grad_norm": 0.4833173453807831,
|
| 288 |
+
"learning_rate": 0.0003819526627218935,
|
| 289 |
+
"loss": 4.554251480102539,
|
| 290 |
+
"step": 400
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 2.42603550295858,
|
| 294 |
+
"grad_norm": 0.6758867502212524,
|
| 295 |
+
"learning_rate": 0.0003789940828402367,
|
| 296 |
+
"loss": 4.56910400390625,
|
| 297 |
+
"step": 410
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 2.485207100591716,
|
| 301 |
+
"grad_norm": 0.6613295078277588,
|
| 302 |
+
"learning_rate": 0.00037603550295857994,
|
| 303 |
+
"loss": 4.526834487915039,
|
| 304 |
+
"step": 420
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 2.544378698224852,
|
| 308 |
+
"grad_norm": 0.5260049104690552,
|
| 309 |
+
"learning_rate": 0.0003730769230769231,
|
| 310 |
+
"loss": 4.503446960449219,
|
| 311 |
+
"step": 430
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 2.603550295857988,
|
| 315 |
+
"grad_norm": 0.60712069272995,
|
| 316 |
+
"learning_rate": 0.00037011834319526627,
|
| 317 |
+
"loss": 4.499971771240235,
|
| 318 |
+
"step": 440
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 2.662721893491124,
|
| 322 |
+
"grad_norm": 0.5932305455207825,
|
| 323 |
+
"learning_rate": 0.00036715976331360946,
|
| 324 |
+
"loss": 4.478979873657226,
|
| 325 |
+
"step": 450
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 2.7218934911242605,
|
| 329 |
+
"grad_norm": 0.6632292866706848,
|
| 330 |
+
"learning_rate": 0.0003642011834319527,
|
| 331 |
+
"loss": 4.521138763427734,
|
| 332 |
+
"step": 460
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 2.7810650887573964,
|
| 336 |
+
"grad_norm": 0.6352905035018921,
|
| 337 |
+
"learning_rate": 0.00036124260355029585,
|
| 338 |
+
"loss": 4.520418930053711,
|
| 339 |
+
"step": 470
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 2.8402366863905324,
|
| 343 |
+
"grad_norm": 0.5262125730514526,
|
| 344 |
+
"learning_rate": 0.00035828402366863904,
|
| 345 |
+
"loss": 4.479636764526367,
|
| 346 |
+
"step": 480
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 2.899408284023669,
|
| 350 |
+
"grad_norm": 0.5989189147949219,
|
| 351 |
+
"learning_rate": 0.0003553254437869823,
|
| 352 |
+
"loss": 4.487276077270508,
|
| 353 |
+
"step": 490
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 2.9585798816568047,
|
| 357 |
+
"grad_norm": 0.767411470413208,
|
| 358 |
+
"learning_rate": 0.0003523668639053255,
|
| 359 |
+
"loss": 4.505224227905273,
|
| 360 |
+
"step": 500
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 3.0177514792899407,
|
| 364 |
+
"grad_norm": 0.5135788321495056,
|
| 365 |
+
"learning_rate": 0.0003494082840236686,
|
| 366 |
+
"loss": 4.422508621215821,
|
| 367 |
+
"step": 510
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 3.076923076923077,
|
| 371 |
+
"grad_norm": 0.6293925046920776,
|
| 372 |
+
"learning_rate": 0.0003464497041420118,
|
| 373 |
+
"loss": 4.3678031921386715,
|
| 374 |
+
"step": 520
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 3.136094674556213,
|
| 378 |
+
"grad_norm": 0.6665500998497009,
|
| 379 |
+
"learning_rate": 0.00034349112426035507,
|
| 380 |
+
"loss": 4.388924789428711,
|
| 381 |
+
"step": 530
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 3.195266272189349,
|
| 385 |
+
"grad_norm": 0.5690398216247559,
|
| 386 |
+
"learning_rate": 0.00034053254437869826,
|
| 387 |
+
"loss": 4.389263153076172,
|
| 388 |
+
"step": 540
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 3.2544378698224854,
|
| 392 |
+
"grad_norm": 0.6139602065086365,
|
| 393 |
+
"learning_rate": 0.0003375739644970414,
|
| 394 |
+
"loss": 4.390654754638672,
|
| 395 |
+
"step": 550
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 3.3136094674556213,
|
| 399 |
+
"grad_norm": 0.5563338398933411,
|
| 400 |
+
"learning_rate": 0.0003346153846153846,
|
| 401 |
+
"loss": 4.337279510498047,
|
| 402 |
+
"step": 560
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 3.3727810650887573,
|
| 406 |
+
"grad_norm": 0.5675852298736572,
|
| 407 |
+
"learning_rate": 0.00033165680473372784,
|
| 408 |
+
"loss": 4.345052337646484,
|
| 409 |
+
"step": 570
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 3.4319526627218933,
|
| 413 |
+
"grad_norm": 0.5776380300521851,
|
| 414 |
+
"learning_rate": 0.00032869822485207103,
|
| 415 |
+
"loss": 4.363264846801758,
|
| 416 |
+
"step": 580
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 3.4911242603550297,
|
| 420 |
+
"grad_norm": 0.6946626901626587,
|
| 421 |
+
"learning_rate": 0.0003257396449704142,
|
| 422 |
+
"loss": 4.355237579345703,
|
| 423 |
+
"step": 590
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 3.5502958579881656,
|
| 427 |
+
"grad_norm": 0.6529305577278137,
|
| 428 |
+
"learning_rate": 0.0003227810650887574,
|
| 429 |
+
"loss": 4.337129211425781,
|
| 430 |
+
"step": 600
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 3.609467455621302,
|
| 434 |
+
"grad_norm": 0.5602537989616394,
|
| 435 |
+
"learning_rate": 0.0003198224852071006,
|
| 436 |
+
"loss": 4.341841125488282,
|
| 437 |
+
"step": 610
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 3.668639053254438,
|
| 441 |
+
"grad_norm": 0.5301386713981628,
|
| 442 |
+
"learning_rate": 0.0003168639053254438,
|
| 443 |
+
"loss": 4.362986755371094,
|
| 444 |
+
"step": 620
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 3.727810650887574,
|
| 448 |
+
"grad_norm": 0.6010480523109436,
|
| 449 |
+
"learning_rate": 0.00031390532544378695,
|
| 450 |
+
"loss": 4.322038269042968,
|
| 451 |
+
"step": 630
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 3.78698224852071,
|
| 455 |
+
"grad_norm": 0.6022279858589172,
|
| 456 |
+
"learning_rate": 0.0003109467455621302,
|
| 457 |
+
"loss": 4.359962844848633,
|
| 458 |
+
"step": 640
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 3.8461538461538463,
|
| 462 |
+
"grad_norm": 0.56232088804245,
|
| 463 |
+
"learning_rate": 0.0003079881656804734,
|
| 464 |
+
"loss": 4.3375293731689455,
|
| 465 |
+
"step": 650
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 3.905325443786982,
|
| 469 |
+
"grad_norm": 0.7015820741653442,
|
| 470 |
+
"learning_rate": 0.0003050295857988166,
|
| 471 |
+
"loss": 4.334150695800782,
|
| 472 |
+
"step": 660
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 3.9644970414201186,
|
| 476 |
+
"grad_norm": 0.6077319383621216,
|
| 477 |
+
"learning_rate": 0.0003020710059171597,
|
| 478 |
+
"loss": 4.341176986694336,
|
| 479 |
+
"step": 670
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 4.023668639053255,
|
| 483 |
+
"grad_norm": 0.773983359336853,
|
| 484 |
+
"learning_rate": 0.00029911242603550297,
|
| 485 |
+
"loss": 4.262639999389648,
|
| 486 |
+
"step": 680
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 4.0828402366863905,
|
| 490 |
+
"grad_norm": 0.5960440635681152,
|
| 491 |
+
"learning_rate": 0.00029615384615384616,
|
| 492 |
+
"loss": 4.2081657409667965,
|
| 493 |
+
"step": 690
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 4.1420118343195265,
|
| 497 |
+
"grad_norm": 0.5495953559875488,
|
| 498 |
+
"learning_rate": 0.00029319526627218936,
|
| 499 |
+
"loss": 4.228216171264648,
|
| 500 |
+
"step": 700
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 4.201183431952662,
|
| 504 |
+
"grad_norm": 0.6150968074798584,
|
| 505 |
+
"learning_rate": 0.0002902366863905325,
|
| 506 |
+
"loss": 4.252474975585938,
|
| 507 |
+
"step": 710
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 4.260355029585799,
|
| 511 |
+
"grad_norm": 0.7338055372238159,
|
| 512 |
+
"learning_rate": 0.00028727810650887574,
|
| 513 |
+
"loss": 4.226696395874024,
|
| 514 |
+
"step": 720
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 4.319526627218935,
|
| 518 |
+
"grad_norm": 0.6750577092170715,
|
| 519 |
+
"learning_rate": 0.00028431952662721894,
|
| 520 |
+
"loss": 4.212493133544922,
|
| 521 |
+
"step": 730
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 4.378698224852071,
|
| 525 |
+
"grad_norm": 0.6345236897468567,
|
| 526 |
+
"learning_rate": 0.00028136094674556213,
|
| 527 |
+
"loss": 4.218838500976562,
|
| 528 |
+
"step": 740
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 4.437869822485207,
|
| 532 |
+
"grad_norm": 0.6776763200759888,
|
| 533 |
+
"learning_rate": 0.0002784023668639054,
|
| 534 |
+
"loss": 4.217085266113282,
|
| 535 |
+
"step": 750
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 4.497041420118343,
|
| 539 |
+
"grad_norm": 0.6580626368522644,
|
| 540 |
+
"learning_rate": 0.0002754437869822485,
|
| 541 |
+
"loss": 4.237897491455078,
|
| 542 |
+
"step": 760
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 4.556213017751479,
|
| 546 |
+
"grad_norm": 0.5971556305885315,
|
| 547 |
+
"learning_rate": 0.0002724852071005917,
|
| 548 |
+
"loss": 4.22449951171875,
|
| 549 |
+
"step": 770
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 4.615384615384615,
|
| 553 |
+
"grad_norm": 0.6069461107254028,
|
| 554 |
+
"learning_rate": 0.0002695266272189349,
|
| 555 |
+
"loss": 4.191236877441407,
|
| 556 |
+
"step": 780
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 4.674556213017752,
|
| 560 |
+
"grad_norm": 0.7110677361488342,
|
| 561 |
+
"learning_rate": 0.00026656804733727815,
|
| 562 |
+
"loss": 4.21960334777832,
|
| 563 |
+
"step": 790
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 4.733727810650888,
|
| 567 |
+
"grad_norm": 0.6254327893257141,
|
| 568 |
+
"learning_rate": 0.0002636094674556213,
|
| 569 |
+
"loss": 4.211502456665039,
|
| 570 |
+
"step": 800
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 4.792899408284024,
|
| 574 |
+
"grad_norm": 0.6190422773361206,
|
| 575 |
+
"learning_rate": 0.0002606508875739645,
|
| 576 |
+
"loss": 4.2044940948486325,
|
| 577 |
+
"step": 810
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 4.85207100591716,
|
| 581 |
+
"grad_norm": 0.6811033487319946,
|
| 582 |
+
"learning_rate": 0.0002576923076923077,
|
| 583 |
+
"loss": 4.237732696533203,
|
| 584 |
+
"step": 820
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 4.911242603550296,
|
| 588 |
+
"grad_norm": 0.5956006646156311,
|
| 589 |
+
"learning_rate": 0.00025473372781065093,
|
| 590 |
+
"loss": 4.215926742553711,
|
| 591 |
+
"step": 830
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 4.970414201183432,
|
| 595 |
+
"grad_norm": 0.6755359172821045,
|
| 596 |
+
"learning_rate": 0.00025177514792899407,
|
| 597 |
+
"loss": 4.18463363647461,
|
| 598 |
+
"step": 840
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 5.029585798816568,
|
| 602 |
+
"grad_norm": 0.6072905659675598,
|
| 603 |
+
"learning_rate": 0.0002488165680473373,
|
| 604 |
+
"loss": 4.154659271240234,
|
| 605 |
+
"step": 850
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 5.088757396449704,
|
| 609 |
+
"grad_norm": 0.6316264271736145,
|
| 610 |
+
"learning_rate": 0.00024585798816568045,
|
| 611 |
+
"loss": 4.110084533691406,
|
| 612 |
+
"step": 860
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 5.14792899408284,
|
| 616 |
+
"grad_norm": 0.5995630621910095,
|
| 617 |
+
"learning_rate": 0.00024289940828402368,
|
| 618 |
+
"loss": 4.094839477539063,
|
| 619 |
+
"step": 870
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 5.207100591715976,
|
| 623 |
+
"grad_norm": 0.5662926435470581,
|
| 624 |
+
"learning_rate": 0.00023994082840236687,
|
| 625 |
+
"loss": 4.120201492309571,
|
| 626 |
+
"step": 880
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 5.266272189349112,
|
| 630 |
+
"grad_norm": 0.5866499543190002,
|
| 631 |
+
"learning_rate": 0.00023698224852071006,
|
| 632 |
+
"loss": 4.091806411743164,
|
| 633 |
+
"step": 890
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 5.325443786982248,
|
| 637 |
+
"grad_norm": 0.6538463830947876,
|
| 638 |
+
"learning_rate": 0.00023402366863905326,
|
| 639 |
+
"loss": 4.110548400878907,
|
| 640 |
+
"step": 900
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 5.384615384615385,
|
| 644 |
+
"grad_norm": 0.7380921840667725,
|
| 645 |
+
"learning_rate": 0.00023106508875739645,
|
| 646 |
+
"loss": 4.089152145385742,
|
| 647 |
+
"step": 910
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 5.443786982248521,
|
| 651 |
+
"grad_norm": 0.7094395756721497,
|
| 652 |
+
"learning_rate": 0.00022810650887573964,
|
| 653 |
+
"loss": 4.125375747680664,
|
| 654 |
+
"step": 920
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 5.502958579881657,
|
| 658 |
+
"grad_norm": 0.6104142069816589,
|
| 659 |
+
"learning_rate": 0.00022514792899408284,
|
| 660 |
+
"loss": 4.108364105224609,
|
| 661 |
+
"step": 930
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 5.562130177514793,
|
| 665 |
+
"grad_norm": 0.6437958478927612,
|
| 666 |
+
"learning_rate": 0.00022218934911242606,
|
| 667 |
+
"loss": 4.102787780761719,
|
| 668 |
+
"step": 940
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 5.621301775147929,
|
| 672 |
+
"grad_norm": 0.6497628092765808,
|
| 673 |
+
"learning_rate": 0.00021923076923076922,
|
| 674 |
+
"loss": 4.098801803588867,
|
| 675 |
+
"step": 950
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 5.680473372781065,
|
| 679 |
+
"grad_norm": 0.7420417666435242,
|
| 680 |
+
"learning_rate": 0.00021627218934911244,
|
| 681 |
+
"loss": 4.110585784912109,
|
| 682 |
+
"step": 960
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 5.739644970414201,
|
| 686 |
+
"grad_norm": 0.6537092924118042,
|
| 687 |
+
"learning_rate": 0.0002133136094674556,
|
| 688 |
+
"loss": 4.079868316650391,
|
| 689 |
+
"step": 970
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 5.798816568047338,
|
| 693 |
+
"grad_norm": 0.9173883199691772,
|
| 694 |
+
"learning_rate": 0.00021035502958579883,
|
| 695 |
+
"loss": 4.0837654113769535,
|
| 696 |
+
"step": 980
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 5.8579881656804735,
|
| 700 |
+
"grad_norm": 0.6213467121124268,
|
| 701 |
+
"learning_rate": 0.000207396449704142,
|
| 702 |
+
"loss": 4.101559829711914,
|
| 703 |
+
"step": 990
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 5.9171597633136095,
|
| 707 |
+
"grad_norm": 0.700158417224884,
|
| 708 |
+
"learning_rate": 0.00020443786982248522,
|
| 709 |
+
"loss": 4.107759475708008,
|
| 710 |
+
"step": 1000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 5.976331360946745,
|
| 714 |
+
"grad_norm": 0.7389625906944275,
|
| 715 |
+
"learning_rate": 0.00020147928994082839,
|
| 716 |
+
"loss": 4.045748901367188,
|
| 717 |
+
"step": 1010
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 6.035502958579881,
|
| 721 |
+
"grad_norm": 0.629220724105835,
|
| 722 |
+
"learning_rate": 0.0001985207100591716,
|
| 723 |
+
"loss": 4.003813552856445,
|
| 724 |
+
"step": 1020
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 6.094674556213017,
|
| 728 |
+
"grad_norm": 0.6587775349617004,
|
| 729 |
+
"learning_rate": 0.00019556213017751477,
|
| 730 |
+
"loss": 3.947599411010742,
|
| 731 |
+
"step": 1030
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 6.153846153846154,
|
| 735 |
+
"grad_norm": 0.6797946691513062,
|
| 736 |
+
"learning_rate": 0.000192603550295858,
|
| 737 |
+
"loss": 3.9570587158203123,
|
| 738 |
+
"step": 1040
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 6.21301775147929,
|
| 742 |
+
"grad_norm": 0.8000279068946838,
|
| 743 |
+
"learning_rate": 0.00018964497041420121,
|
| 744 |
+
"loss": 3.9946006774902343,
|
| 745 |
+
"step": 1050
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 6.272189349112426,
|
| 749 |
+
"grad_norm": 0.7699520587921143,
|
| 750 |
+
"learning_rate": 0.00018668639053254438,
|
| 751 |
+
"loss": 3.9914344787597655,
|
| 752 |
+
"step": 1060
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 6.331360946745562,
|
| 756 |
+
"grad_norm": 0.7474783062934875,
|
| 757 |
+
"learning_rate": 0.0001837278106508876,
|
| 758 |
+
"loss": 3.9896411895751953,
|
| 759 |
+
"step": 1070
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 6.390532544378698,
|
| 763 |
+
"grad_norm": 0.6568964719772339,
|
| 764 |
+
"learning_rate": 0.00018076923076923077,
|
| 765 |
+
"loss": 3.9945484161376954,
|
| 766 |
+
"step": 1080
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 6.449704142011834,
|
| 770 |
+
"grad_norm": 0.702962338924408,
|
| 771 |
+
"learning_rate": 0.000177810650887574,
|
| 772 |
+
"loss": 3.9896453857421874,
|
| 773 |
+
"step": 1090
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 6.508875739644971,
|
| 777 |
+
"grad_norm": 0.8072907328605652,
|
| 778 |
+
"learning_rate": 0.00017485207100591716,
|
| 779 |
+
"loss": 3.9789073944091795,
|
| 780 |
+
"step": 1100
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 6.568047337278107,
|
| 784 |
+
"grad_norm": 0.690667986869812,
|
| 785 |
+
"learning_rate": 0.00017189349112426038,
|
| 786 |
+
"loss": 3.9628318786621093,
|
| 787 |
+
"step": 1110
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 6.627218934911243,
|
| 791 |
+
"grad_norm": 0.7332726716995239,
|
| 792 |
+
"learning_rate": 0.00016893491124260354,
|
| 793 |
+
"loss": 4.001555633544922,
|
| 794 |
+
"step": 1120
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 6.686390532544379,
|
| 798 |
+
"grad_norm": 0.7399805784225464,
|
| 799 |
+
"learning_rate": 0.00016597633136094676,
|
| 800 |
+
"loss": 3.987625503540039,
|
| 801 |
+
"step": 1130
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 6.745562130177515,
|
| 805 |
+
"grad_norm": 1.0256022214889526,
|
| 806 |
+
"learning_rate": 0.00016301775147928993,
|
| 807 |
+
"loss": 3.9571834564208985,
|
| 808 |
+
"step": 1140
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 6.804733727810651,
|
| 812 |
+
"grad_norm": 0.7827850580215454,
|
| 813 |
+
"learning_rate": 0.00016005917159763315,
|
| 814 |
+
"loss": 3.9762195587158202,
|
| 815 |
+
"step": 1150
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 6.8639053254437865,
|
| 819 |
+
"grad_norm": 0.7092675566673279,
|
| 820 |
+
"learning_rate": 0.00015710059171597632,
|
| 821 |
+
"loss": 3.9586578369140626,
|
| 822 |
+
"step": 1160
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 6.923076923076923,
|
| 826 |
+
"grad_norm": 0.7039374709129333,
|
| 827 |
+
"learning_rate": 0.00015414201183431954,
|
| 828 |
+
"loss": 3.944875717163086,
|
| 829 |
+
"step": 1170
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 6.982248520710059,
|
| 833 |
+
"grad_norm": 0.7573893070220947,
|
| 834 |
+
"learning_rate": 0.00015118343195266273,
|
| 835 |
+
"loss": 3.975397491455078,
|
| 836 |
+
"step": 1180
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 7.041420118343195,
|
| 840 |
+
"grad_norm": 0.7940344214439392,
|
| 841 |
+
"learning_rate": 0.00014822485207100592,
|
| 842 |
+
"loss": 3.8859161376953124,
|
| 843 |
+
"step": 1190
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 7.100591715976331,
|
| 847 |
+
"grad_norm": 0.8005825877189636,
|
| 848 |
+
"learning_rate": 0.00014526627218934912,
|
| 849 |
+
"loss": 3.8561363220214844,
|
| 850 |
+
"step": 1200
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 7.159763313609467,
|
| 854 |
+
"grad_norm": 0.7510560750961304,
|
| 855 |
+
"learning_rate": 0.0001423076923076923,
|
| 856 |
+
"loss": 3.842996597290039,
|
| 857 |
+
"step": 1210
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 7.218934911242604,
|
| 861 |
+
"grad_norm": 0.7474063038825989,
|
| 862 |
+
"learning_rate": 0.0001393491124260355,
|
| 863 |
+
"loss": 3.8655860900878904,
|
| 864 |
+
"step": 1220
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 7.27810650887574,
|
| 868 |
+
"grad_norm": 0.7399153113365173,
|
| 869 |
+
"learning_rate": 0.0001363905325443787,
|
| 870 |
+
"loss": 3.8691051483154295,
|
| 871 |
+
"step": 1230
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 7.337278106508876,
|
| 875 |
+
"grad_norm": 0.6961866021156311,
|
| 876 |
+
"learning_rate": 0.0001334319526627219,
|
| 877 |
+
"loss": 3.8814228057861326,
|
| 878 |
+
"step": 1240
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 7.396449704142012,
|
| 882 |
+
"grad_norm": 0.7503266334533691,
|
| 883 |
+
"learning_rate": 0.00013047337278106509,
|
| 884 |
+
"loss": 3.860675811767578,
|
| 885 |
+
"step": 1250
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 7.455621301775148,
|
| 889 |
+
"grad_norm": 0.7707279920578003,
|
| 890 |
+
"learning_rate": 0.00012751479289940828,
|
| 891 |
+
"loss": 3.8663909912109373,
|
| 892 |
+
"step": 1260
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 7.514792899408284,
|
| 896 |
+
"grad_norm": 0.7248626947402954,
|
| 897 |
+
"learning_rate": 0.00012455621301775147,
|
| 898 |
+
"loss": 3.8687854766845704,
|
| 899 |
+
"step": 1270
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 7.57396449704142,
|
| 903 |
+
"grad_norm": 0.7378124594688416,
|
| 904 |
+
"learning_rate": 0.00012159763313609468,
|
| 905 |
+
"loss": 3.853357696533203,
|
| 906 |
+
"step": 1280
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 7.633136094674557,
|
| 910 |
+
"grad_norm": 0.7708460092544556,
|
| 911 |
+
"learning_rate": 0.00011863905325443787,
|
| 912 |
+
"loss": 3.8905601501464844,
|
| 913 |
+
"step": 1290
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 7.6923076923076925,
|
| 917 |
+
"grad_norm": 0.6937795877456665,
|
| 918 |
+
"learning_rate": 0.00011568047337278107,
|
| 919 |
+
"loss": 3.87835578918457,
|
| 920 |
+
"step": 1300
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 7.7514792899408285,
|
| 924 |
+
"grad_norm": 0.8977406620979309,
|
| 925 |
+
"learning_rate": 0.00011272189349112426,
|
| 926 |
+
"loss": 3.865726852416992,
|
| 927 |
+
"step": 1310
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 7.810650887573964,
|
| 931 |
+
"grad_norm": 0.7240127921104431,
|
| 932 |
+
"learning_rate": 0.00010976331360946746,
|
| 933 |
+
"loss": 3.846134567260742,
|
| 934 |
+
"step": 1320
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 7.8698224852071,
|
| 938 |
+
"grad_norm": 0.7820615768432617,
|
| 939 |
+
"learning_rate": 0.00010680473372781065,
|
| 940 |
+
"loss": 3.8235801696777343,
|
| 941 |
+
"step": 1330
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 7.928994082840236,
|
| 945 |
+
"grad_norm": 0.7931578755378723,
|
| 946 |
+
"learning_rate": 0.00010384615384615386,
|
| 947 |
+
"loss": 3.852934646606445,
|
| 948 |
+
"step": 1340
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 7.988165680473373,
|
| 952 |
+
"grad_norm": 0.7435231804847717,
|
| 953 |
+
"learning_rate": 0.00010088757396449705,
|
| 954 |
+
"loss": 3.8489166259765626,
|
| 955 |
+
"step": 1350
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 8.04733727810651,
|
| 959 |
+
"grad_norm": 0.7691823244094849,
|
| 960 |
+
"learning_rate": 9.792899408284024e-05,
|
| 961 |
+
"loss": 3.794635009765625,
|
| 962 |
+
"step": 1360
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 8.106508875739644,
|
| 966 |
+
"grad_norm": 0.885684609413147,
|
| 967 |
+
"learning_rate": 9.497041420118344e-05,
|
| 968 |
+
"loss": 3.787654495239258,
|
| 969 |
+
"step": 1370
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 8.165680473372781,
|
| 973 |
+
"grad_norm": 0.8940234184265137,
|
| 974 |
+
"learning_rate": 9.201183431952663e-05,
|
| 975 |
+
"loss": 3.756052780151367,
|
| 976 |
+
"step": 1380
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 8.224852071005918,
|
| 980 |
+
"grad_norm": 0.7225572466850281,
|
| 981 |
+
"learning_rate": 8.905325443786982e-05,
|
| 982 |
+
"loss": 3.724110412597656,
|
| 983 |
+
"step": 1390
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 8.284023668639053,
|
| 987 |
+
"grad_norm": 0.792807936668396,
|
| 988 |
+
"learning_rate": 8.609467455621302e-05,
|
| 989 |
+
"loss": 3.75240478515625,
|
| 990 |
+
"step": 1400
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 8.34319526627219,
|
| 994 |
+
"grad_norm": 0.8160039782524109,
|
| 995 |
+
"learning_rate": 8.313609467455621e-05,
|
| 996 |
+
"loss": 3.7542896270751953,
|
| 997 |
+
"step": 1410
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 8.402366863905325,
|
| 1001 |
+
"grad_norm": 0.715473473072052,
|
| 1002 |
+
"learning_rate": 8.01775147928994e-05,
|
| 1003 |
+
"loss": 3.817658233642578,
|
| 1004 |
+
"step": 1420
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 8.461538461538462,
|
| 1008 |
+
"grad_norm": 0.8090062141418457,
|
| 1009 |
+
"learning_rate": 7.72189349112426e-05,
|
| 1010 |
+
"loss": 3.7604396820068358,
|
| 1011 |
+
"step": 1430
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 8.520710059171599,
|
| 1015 |
+
"grad_norm": 0.726865828037262,
|
| 1016 |
+
"learning_rate": 7.426035502958579e-05,
|
| 1017 |
+
"loss": 3.7455059051513673,
|
| 1018 |
+
"step": 1440
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 8.579881656804734,
|
| 1022 |
+
"grad_norm": 0.7755900621414185,
|
| 1023 |
+
"learning_rate": 7.1301775147929e-05,
|
| 1024 |
+
"loss": 3.7822811126708986,
|
| 1025 |
+
"step": 1450
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 8.63905325443787,
|
| 1029 |
+
"grad_norm": 0.7525601983070374,
|
| 1030 |
+
"learning_rate": 6.834319526627219e-05,
|
| 1031 |
+
"loss": 3.7710426330566404,
|
| 1032 |
+
"step": 1460
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 8.698224852071005,
|
| 1036 |
+
"grad_norm": 0.7370349764823914,
|
| 1037 |
+
"learning_rate": 6.538461538461539e-05,
|
| 1038 |
+
"loss": 3.7621829986572264,
|
| 1039 |
+
"step": 1470
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 8.757396449704142,
|
| 1043 |
+
"grad_norm": 0.7664944529533386,
|
| 1044 |
+
"learning_rate": 6.242603550295858e-05,
|
| 1045 |
+
"loss": 3.7525020599365235,
|
| 1046 |
+
"step": 1480
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 8.816568047337277,
|
| 1050 |
+
"grad_norm": 0.8167346119880676,
|
| 1051 |
+
"learning_rate": 5.946745562130178e-05,
|
| 1052 |
+
"loss": 3.7805561065673827,
|
| 1053 |
+
"step": 1490
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 8.875739644970414,
|
| 1057 |
+
"grad_norm": 0.7823934555053711,
|
| 1058 |
+
"learning_rate": 5.6508875739644974e-05,
|
| 1059 |
+
"loss": 3.766466522216797,
|
| 1060 |
+
"step": 1500
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 8.934911242603551,
|
| 1064 |
+
"grad_norm": 0.8203352093696594,
|
| 1065 |
+
"learning_rate": 5.355029585798817e-05,
|
| 1066 |
+
"loss": 3.7743297576904298,
|
| 1067 |
+
"step": 1510
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 8.994082840236686,
|
| 1071 |
+
"grad_norm": 0.7480125427246094,
|
| 1072 |
+
"learning_rate": 5.059171597633136e-05,
|
| 1073 |
+
"loss": 3.7630767822265625,
|
| 1074 |
+
"step": 1520
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 9.053254437869823,
|
| 1078 |
+
"grad_norm": 0.7367929220199585,
|
| 1079 |
+
"learning_rate": 4.763313609467456e-05,
|
| 1080 |
+
"loss": 3.7045642852783205,
|
| 1081 |
+
"step": 1530
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 9.112426035502958,
|
| 1085 |
+
"grad_norm": 0.7291252017021179,
|
| 1086 |
+
"learning_rate": 4.4674556213017755e-05,
|
| 1087 |
+
"loss": 3.7382125854492188,
|
| 1088 |
+
"step": 1540
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 9.171597633136095,
|
| 1092 |
+
"grad_norm": 0.7637320756912231,
|
| 1093 |
+
"learning_rate": 4.171597633136095e-05,
|
| 1094 |
+
"loss": 3.7042491912841795,
|
| 1095 |
+
"step": 1550
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 9.23076923076923,
|
| 1099 |
+
"grad_norm": 0.7738555669784546,
|
| 1100 |
+
"learning_rate": 3.875739644970414e-05,
|
| 1101 |
+
"loss": 3.703077697753906,
|
| 1102 |
+
"step": 1560
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 9.289940828402367,
|
| 1106 |
+
"grad_norm": 0.7595531940460205,
|
| 1107 |
+
"learning_rate": 3.5798816568047336e-05,
|
| 1108 |
+
"loss": 3.7166164398193358,
|
| 1109 |
+
"step": 1570
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 9.349112426035504,
|
| 1113 |
+
"grad_norm": 0.7209826707839966,
|
| 1114 |
+
"learning_rate": 3.284023668639053e-05,
|
| 1115 |
+
"loss": 3.7272975921630858,
|
| 1116 |
+
"step": 1580
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 9.408284023668639,
|
| 1120 |
+
"grad_norm": 0.7385886907577515,
|
| 1121 |
+
"learning_rate": 2.9881656804733726e-05,
|
| 1122 |
+
"loss": 3.710719680786133,
|
| 1123 |
+
"step": 1590
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 9.467455621301776,
|
| 1127 |
+
"grad_norm": 0.7908353805541992,
|
| 1128 |
+
"learning_rate": 2.6923076923076927e-05,
|
| 1129 |
+
"loss": 3.7108623504638674,
|
| 1130 |
+
"step": 1600
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 9.52662721893491,
|
| 1134 |
+
"grad_norm": 0.7293062210083008,
|
| 1135 |
+
"learning_rate": 2.396449704142012e-05,
|
| 1136 |
+
"loss": 3.6832290649414063,
|
| 1137 |
+
"step": 1610
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 9.585798816568047,
|
| 1141 |
+
"grad_norm": 0.7990091443061829,
|
| 1142 |
+
"learning_rate": 2.1005917159763314e-05,
|
| 1143 |
+
"loss": 3.693344497680664,
|
| 1144 |
+
"step": 1620
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 9.644970414201183,
|
| 1148 |
+
"grad_norm": 0.7727529406547546,
|
| 1149 |
+
"learning_rate": 1.804733727810651e-05,
|
| 1150 |
+
"loss": 3.704065704345703,
|
| 1151 |
+
"step": 1630
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 9.70414201183432,
|
| 1155 |
+
"grad_norm": 0.7558280825614929,
|
| 1156 |
+
"learning_rate": 1.5088757396449705e-05,
|
| 1157 |
+
"loss": 3.701119232177734,
|
| 1158 |
+
"step": 1640
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 9.763313609467456,
|
| 1162 |
+
"grad_norm": 0.7217801213264465,
|
| 1163 |
+
"learning_rate": 1.21301775147929e-05,
|
| 1164 |
+
"loss": 3.6759807586669924,
|
| 1165 |
+
"step": 1650
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 9.822485207100591,
|
| 1169 |
+
"grad_norm": 0.7166690826416016,
|
| 1170 |
+
"learning_rate": 9.171597633136093e-06,
|
| 1171 |
+
"loss": 3.691928482055664,
|
| 1172 |
+
"step": 1660
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 9.881656804733728,
|
| 1176 |
+
"grad_norm": 0.7086900472640991,
|
| 1177 |
+
"learning_rate": 6.2130177514792895e-06,
|
| 1178 |
+
"loss": 3.679682159423828,
|
| 1179 |
+
"step": 1670
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 9.940828402366863,
|
| 1183 |
+
"grad_norm": 0.7332860827445984,
|
| 1184 |
+
"learning_rate": 3.2544378698224853e-06,
|
| 1185 |
+
"loss": 3.7041976928710936,
|
| 1186 |
+
"step": 1680
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 10.0,
|
| 1190 |
+
"grad_norm": 0.7306056022644043,
|
| 1191 |
+
"learning_rate": 2.958579881656805e-07,
|
| 1192 |
+
"loss": 3.6652008056640626,
|
| 1193 |
+
"step": 1690
|
| 1194 |
+
}
|
| 1195 |
+
],
|
| 1196 |
+
"logging_steps": 10,
|
| 1197 |
+
"max_steps": 1690,
|
| 1198 |
+
"num_input_tokens_seen": 0,
|
| 1199 |
+
"num_train_epochs": 10,
|
| 1200 |
+
"save_steps": 999999,
|
| 1201 |
+
"stateful_callbacks": {
|
| 1202 |
+
"TrainerControl": {
|
| 1203 |
+
"args": {
|
| 1204 |
+
"should_epoch_stop": false,
|
| 1205 |
+
"should_evaluate": false,
|
| 1206 |
+
"should_log": false,
|
| 1207 |
+
"should_save": true,
|
| 1208 |
+
"should_training_stop": true
|
| 1209 |
+
},
|
| 1210 |
+
"attributes": {}
|
| 1211 |
+
}
|
| 1212 |
+
},
|
| 1213 |
+
"total_flos": 1047519874252800.0,
|
| 1214 |
+
"train_batch_size": 8,
|
| 1215 |
+
"trial_name": null,
|
| 1216 |
+
"trial_params": null
|
| 1217 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a7c7d7fcaea27cb058c2217cb853f72abca049f81a8668821cc08d49b5f85ea4
|
| 3 |
+
size 5137
|