Upload folder using huggingface_hub
Browse files
slots/8/checkpoint-6037/config.json
ADDED
|
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"SsaiForCausalLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_dropout": 0.0,
|
| 6 |
+
"dtype": "float32",
|
| 7 |
+
"hidden_dropout": 0.0,
|
| 8 |
+
"hidden_size": 768,
|
| 9 |
+
"initializer_range": 0.02,
|
| 10 |
+
"intermediate_size": 3072,
|
| 11 |
+
"max_position_embeddings": 128,
|
| 12 |
+
"model_type": "ssai",
|
| 13 |
+
"neftune_alpha": 0.0,
|
| 14 |
+
"num_attention_heads": 12,
|
| 15 |
+
"num_hidden_layers": 13,
|
| 16 |
+
"num_key_value_heads": 3,
|
| 17 |
+
"qk_norm": true,
|
| 18 |
+
"rezero_init": 1.0,
|
| 19 |
+
"rms_norm_eps": 1e-06,
|
| 20 |
+
"rope_theta": 10000.0,
|
| 21 |
+
"transformers_version": "5.0.0",
|
| 22 |
+
"use_cache": false,
|
| 23 |
+
"vocab_size": 32000
|
| 24 |
+
}
|
slots/8/checkpoint-6037/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:99da0af6fc95c2a79641fd2fcac80044ddbb512c0f3b3c7aa47f93859e4c77ac
|
| 3 |
+
size 672310484
|
slots/8/checkpoint-6037/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8ae7f1f05e583f2b0a776aa748b7fa0518089c8559d8d3588f5bd113284ed973
|
| 3 |
+
size 562589195
|
slots/8/checkpoint-6037/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5ad0bfd3a2229f075c7d673864e2566c3fc4474dc92e8caa0e2385a8fd65cfb9
|
| 3 |
+
size 1383
|
slots/8/checkpoint-6037/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6cca684d91cd4af9d33fccba9103e933db8821ad6337d643505438c7d3c0a22d
|
| 3 |
+
size 1465
|
slots/8/checkpoint-6037/trainer_state.json
ADDED
|
@@ -0,0 +1,2148 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.1279702102365216,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 6037,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 2.1197654214590477e-05,
|
| 14 |
+
"grad_norm": 65.25238800048828,
|
| 15 |
+
"learning_rate": 5e-05,
|
| 16 |
+
"loss": 221.60691833496094,
|
| 17 |
+
"step": 1
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.00042395308429180956,
|
| 21 |
+
"grad_norm": 74.57616424560547,
|
| 22 |
+
"learning_rate": 4.999999853385662e-05,
|
| 23 |
+
"loss": 210.67147666529604,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.0008479061685836191,
|
| 28 |
+
"grad_norm": Infinity,
|
| 29 |
+
"learning_rate": 4.999999346570685e-05,
|
| 30 |
+
"loss": 193.43048095703125,
|
| 31 |
+
"step": 40
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.0012718592528754287,
|
| 35 |
+
"grad_norm": 220.3378143310547,
|
| 36 |
+
"learning_rate": 4.9999985809181255e-05,
|
| 37 |
+
"loss": 178.9335205078125,
|
| 38 |
+
"step": 60
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.0016958123371672382,
|
| 42 |
+
"grad_norm": 256.5308837890625,
|
| 43 |
+
"learning_rate": 4.999997454612581e-05,
|
| 44 |
+
"loss": 178.64384765625,
|
| 45 |
+
"step": 80
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.002119765421459048,
|
| 49 |
+
"grad_norm": 248.0841064453125,
|
| 50 |
+
"learning_rate": 4.999995916067716e-05,
|
| 51 |
+
"loss": 170.77950439453124,
|
| 52 |
+
"step": 100
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.0025437185057508574,
|
| 56 |
+
"grad_norm": 90.63424682617188,
|
| 57 |
+
"learning_rate": 4.999994119138215e-05,
|
| 58 |
+
"loss": 155.6619140625,
|
| 59 |
+
"step": 120
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.002967671590042667,
|
| 63 |
+
"grad_norm": 33.57120132446289,
|
| 64 |
+
"learning_rate": 4.9999918746738674e-05,
|
| 65 |
+
"loss": 148.322412109375,
|
| 66 |
+
"step": 140
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.0033916246743344765,
|
| 70 |
+
"grad_norm": 35.397247314453125,
|
| 71 |
+
"learning_rate": 4.999989268199985e-05,
|
| 72 |
+
"loss": 146.0496826171875,
|
| 73 |
+
"step": 160
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.003815577758626286,
|
| 77 |
+
"grad_norm": 34.25936508178711,
|
| 78 |
+
"learning_rate": 4.9999862997169434e-05,
|
| 79 |
+
"loss": 132.1153076171875,
|
| 80 |
+
"step": 180
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.004239530842918096,
|
| 84 |
+
"grad_norm": 30.451223373413086,
|
| 85 |
+
"learning_rate": 4.9999829692251745e-05,
|
| 86 |
+
"loss": 126.0611328125,
|
| 87 |
+
"step": 200
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.004663483927209905,
|
| 91 |
+
"grad_norm": 23.14311981201172,
|
| 92 |
+
"learning_rate": 4.9999792767251585e-05,
|
| 93 |
+
"loss": 120.37777099609374,
|
| 94 |
+
"step": 220
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.005087437011501715,
|
| 98 |
+
"grad_norm": 31.511438369750977,
|
| 99 |
+
"learning_rate": 4.9999752222174313e-05,
|
| 100 |
+
"loss": 117.167626953125,
|
| 101 |
+
"step": 240
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.005511390095793524,
|
| 105 |
+
"grad_norm": 42.21919631958008,
|
| 106 |
+
"learning_rate": 4.99997080570258e-05,
|
| 107 |
+
"loss": 112.9516845703125,
|
| 108 |
+
"step": 260
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.005935343180085334,
|
| 112 |
+
"grad_norm": 38.3533821105957,
|
| 113 |
+
"learning_rate": 4.9999660271812445e-05,
|
| 114 |
+
"loss": 106.51890869140625,
|
| 115 |
+
"step": 280
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.006359296264377143,
|
| 119 |
+
"grad_norm": 33.42371368408203,
|
| 120 |
+
"learning_rate": 4.999960886654115e-05,
|
| 121 |
+
"loss": 106.55684814453124,
|
| 122 |
+
"step": 300
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.006783249348668953,
|
| 126 |
+
"grad_norm": 43.83552169799805,
|
| 127 |
+
"learning_rate": 4.9999553841219373e-05,
|
| 128 |
+
"loss": 105.79185791015625,
|
| 129 |
+
"step": 320
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.007207202432960762,
|
| 133 |
+
"grad_norm": 40.87639617919922,
|
| 134 |
+
"learning_rate": 4.9999495195855086e-05,
|
| 135 |
+
"loss": 100.90576171875,
|
| 136 |
+
"step": 340
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.007631155517252572,
|
| 140 |
+
"grad_norm": 43.487091064453125,
|
| 141 |
+
"learning_rate": 4.999943293045677e-05,
|
| 142 |
+
"loss": 99.137158203125,
|
| 143 |
+
"step": 360
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.008055108601544382,
|
| 147 |
+
"grad_norm": 39.68036651611328,
|
| 148 |
+
"learning_rate": 4.999936704503345e-05,
|
| 149 |
+
"loss": 95.90462036132813,
|
| 150 |
+
"step": 380
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.008479061685836191,
|
| 154 |
+
"grad_norm": 47.12007141113281,
|
| 155 |
+
"learning_rate": 4.999929753959466e-05,
|
| 156 |
+
"loss": 104.62547607421875,
|
| 157 |
+
"step": 400
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.008903014770128,
|
| 161 |
+
"grad_norm": 40.647132873535156,
|
| 162 |
+
"learning_rate": 4.999922441415046e-05,
|
| 163 |
+
"loss": 92.67885131835938,
|
| 164 |
+
"step": 420
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.00932696785441981,
|
| 168 |
+
"grad_norm": 43.75863265991211,
|
| 169 |
+
"learning_rate": 4.9999147668711454e-05,
|
| 170 |
+
"loss": 92.218115234375,
|
| 171 |
+
"step": 440
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.00975092093871162,
|
| 175 |
+
"grad_norm": 40.93122100830078,
|
| 176 |
+
"learning_rate": 4.999906730328874e-05,
|
| 177 |
+
"loss": 91.65263671875,
|
| 178 |
+
"step": 460
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.01017487402300343,
|
| 182 |
+
"grad_norm": 54.96118927001953,
|
| 183 |
+
"learning_rate": 4.999898331789397e-05,
|
| 184 |
+
"loss": 86.0513427734375,
|
| 185 |
+
"step": 480
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.010598827107295239,
|
| 189 |
+
"grad_norm": 47.806495666503906,
|
| 190 |
+
"learning_rate": 4.9998895712539294e-05,
|
| 191 |
+
"loss": 85.7255859375,
|
| 192 |
+
"step": 500
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.007654710720017547,
|
| 196 |
+
"grad_norm": 965.6040649414062,
|
| 197 |
+
"learning_rate": 4.999999997326698e-05,
|
| 198 |
+
"loss": 98.29871622721355,
|
| 199 |
+
"step": 520
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.007949122670787453,
|
| 203 |
+
"grad_norm": 121.55615234375,
|
| 204 |
+
"learning_rate": 4.999999933167428e-05,
|
| 205 |
+
"loss": 41.383172607421876,
|
| 206 |
+
"step": 540
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.008243534621557358,
|
| 210 |
+
"grad_norm": 168.19332885742188,
|
| 211 |
+
"learning_rate": 4.999999783462468e-05,
|
| 212 |
+
"loss": 34.619232177734375,
|
| 213 |
+
"step": 560
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.008537946572327263,
|
| 217 |
+
"grad_norm": 123.20977783203125,
|
| 218 |
+
"learning_rate": 4.9999995482118227e-05,
|
| 219 |
+
"loss": 32.92503662109375,
|
| 220 |
+
"step": 580
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.006359296264377143,
|
| 224 |
+
"grad_norm": 21.918010711669922,
|
| 225 |
+
"learning_rate": 4.999999956540614e-05,
|
| 226 |
+
"loss": 44.29385555491728,
|
| 227 |
+
"step": 600
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.006571272806523048,
|
| 231 |
+
"grad_norm": 19.172475814819336,
|
| 232 |
+
"learning_rate": 4.9999997436783165e-05,
|
| 233 |
+
"loss": 45.86236572265625,
|
| 234 |
+
"step": 620
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.006783249348668953,
|
| 238 |
+
"grad_norm": 18.668466567993164,
|
| 239 |
+
"learning_rate": 4.999999353430788e-05,
|
| 240 |
+
"loss": 44.35866088867188,
|
| 241 |
+
"step": 640
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.0069952258908148575,
|
| 245 |
+
"grad_norm": 21.126758575439453,
|
| 246 |
+
"learning_rate": 4.9999988183925896e-05,
|
| 247 |
+
"loss": 43.077456665039065,
|
| 248 |
+
"step": 660
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.007207202432960762,
|
| 252 |
+
"grad_norm": 19.767131805419922,
|
| 253 |
+
"learning_rate": 4.999998082243949e-05,
|
| 254 |
+
"loss": 41.443548583984374,
|
| 255 |
+
"step": 680
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.0074191789751066675,
|
| 259 |
+
"grad_norm": 20.20482635498047,
|
| 260 |
+
"learning_rate": 4.999997168710195e-05,
|
| 261 |
+
"loss": 41.82533264160156,
|
| 262 |
+
"step": 700
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.007631155517252572,
|
| 266 |
+
"grad_norm": 21.61404800415039,
|
| 267 |
+
"learning_rate": 4.999996077791391e-05,
|
| 268 |
+
"loss": 40.30120849609375,
|
| 269 |
+
"step": 720
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.007843132059398477,
|
| 273 |
+
"grad_norm": 18.67470932006836,
|
| 274 |
+
"learning_rate": 4.999994809487617e-05,
|
| 275 |
+
"loss": 38.5742431640625,
|
| 276 |
+
"step": 740
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.008055108601544382,
|
| 280 |
+
"grad_norm": 17.836240768432617,
|
| 281 |
+
"learning_rate": 4.999993363798959e-05,
|
| 282 |
+
"loss": 39.95260009765625,
|
| 283 |
+
"step": 760
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.008267085143690286,
|
| 287 |
+
"grad_norm": 21.8165225982666,
|
| 288 |
+
"learning_rate": 4.999991740725524e-05,
|
| 289 |
+
"loss": 40.794134521484374,
|
| 290 |
+
"step": 780
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.008479061685836191,
|
| 294 |
+
"grad_norm": 19.97357940673828,
|
| 295 |
+
"learning_rate": 4.999989940267423e-05,
|
| 296 |
+
"loss": 39.2896240234375,
|
| 297 |
+
"step": 800
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.008691038227982097,
|
| 301 |
+
"grad_norm": 19.85247802734375,
|
| 302 |
+
"learning_rate": 4.999987962424787e-05,
|
| 303 |
+
"loss": 36.8283447265625,
|
| 304 |
+
"step": 820
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.008903014770128,
|
| 308 |
+
"grad_norm": 19.100263595581055,
|
| 309 |
+
"learning_rate": 4.999985807197755e-05,
|
| 310 |
+
"loss": 36.863983154296875,
|
| 311 |
+
"step": 840
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.009114991312273906,
|
| 315 |
+
"grad_norm": 23.29732322692871,
|
| 316 |
+
"learning_rate": 4.999983474586479e-05,
|
| 317 |
+
"loss": 36.45862731933594,
|
| 318 |
+
"step": 860
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.00932696785441981,
|
| 322 |
+
"grad_norm": 15.870063781738281,
|
| 323 |
+
"learning_rate": 4.9999809645911255e-05,
|
| 324 |
+
"loss": 36.08320617675781,
|
| 325 |
+
"step": 880
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.009538944396565715,
|
| 329 |
+
"grad_norm": 15.129697799682617,
|
| 330 |
+
"learning_rate": 4.999978277211873e-05,
|
| 331 |
+
"loss": 34.60118713378906,
|
| 332 |
+
"step": 900
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.00975092093871162,
|
| 336 |
+
"grad_norm": 15.188931465148926,
|
| 337 |
+
"learning_rate": 4.999975412448912e-05,
|
| 338 |
+
"loss": 34.275796508789064,
|
| 339 |
+
"step": 920
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.009962897480857524,
|
| 343 |
+
"grad_norm": 12.236566543579102,
|
| 344 |
+
"learning_rate": 4.999972370302445e-05,
|
| 345 |
+
"loss": 33.634912109375,
|
| 346 |
+
"step": 940
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.01017487402300343,
|
| 350 |
+
"grad_norm": 17.674135208129883,
|
| 351 |
+
"learning_rate": 4.999969150772689e-05,
|
| 352 |
+
"loss": 32.5128662109375,
|
| 353 |
+
"step": 960
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.010386850565149335,
|
| 357 |
+
"grad_norm": 16.264986038208008,
|
| 358 |
+
"learning_rate": 4.999965753859871e-05,
|
| 359 |
+
"loss": 32.47880249023437,
|
| 360 |
+
"step": 980
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.010598827107295239,
|
| 364 |
+
"grad_norm": 15.638131141662598,
|
| 365 |
+
"learning_rate": 4.9999621795642326e-05,
|
| 366 |
+
"loss": 32.50233154296875,
|
| 367 |
+
"step": 1000
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.010810803649441144,
|
| 371 |
+
"grad_norm": 12.960165023803711,
|
| 372 |
+
"learning_rate": 4.999958427886029e-05,
|
| 373 |
+
"loss": 30.927606201171876,
|
| 374 |
+
"step": 1020
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.011022780191587048,
|
| 378 |
+
"grad_norm": 15.181825637817383,
|
| 379 |
+
"learning_rate": 4.9999544988255235e-05,
|
| 380 |
+
"loss": 31.094097900390626,
|
| 381 |
+
"step": 1040
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.011234756733732953,
|
| 385 |
+
"grad_norm": 12.673530578613281,
|
| 386 |
+
"learning_rate": 4.999950392382997e-05,
|
| 387 |
+
"loss": 29.733123779296875,
|
| 388 |
+
"step": 1060
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.011446733275878859,
|
| 392 |
+
"grad_norm": 16.327787399291992,
|
| 393 |
+
"learning_rate": 4.999946108558741e-05,
|
| 394 |
+
"loss": 31.226516723632812,
|
| 395 |
+
"step": 1080
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.011658709818024762,
|
| 399 |
+
"grad_norm": 16.461864471435547,
|
| 400 |
+
"learning_rate": 4.999941647353058e-05,
|
| 401 |
+
"loss": 30.790750122070314,
|
| 402 |
+
"step": 1100
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.011870686360170668,
|
| 406 |
+
"grad_norm": 10.78894329071045,
|
| 407 |
+
"learning_rate": 4.999937008766266e-05,
|
| 408 |
+
"loss": 31.257119750976564,
|
| 409 |
+
"step": 1120
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.012082662902316573,
|
| 413 |
+
"grad_norm": 12.983068466186523,
|
| 414 |
+
"learning_rate": 4.999932192798692e-05,
|
| 415 |
+
"loss": 28.610498046875,
|
| 416 |
+
"step": 1140
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.012294639444462477,
|
| 420 |
+
"grad_norm": 12.125755310058594,
|
| 421 |
+
"learning_rate": 4.99992719945068e-05,
|
| 422 |
+
"loss": 28.451394653320314,
|
| 423 |
+
"step": 1160
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.012506615986608382,
|
| 427 |
+
"grad_norm": 17.49968910217285,
|
| 428 |
+
"learning_rate": 4.999922028722583e-05,
|
| 429 |
+
"loss": 28.062771606445313,
|
| 430 |
+
"step": 1180
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.012718592528754286,
|
| 434 |
+
"grad_norm": 12.746994972229004,
|
| 435 |
+
"learning_rate": 4.999916680614769e-05,
|
| 436 |
+
"loss": 28.245272827148437,
|
| 437 |
+
"step": 1200
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.012930569070900191,
|
| 441 |
+
"grad_norm": 15.722673416137695,
|
| 442 |
+
"learning_rate": 4.9999111551276165e-05,
|
| 443 |
+
"loss": 27.170697021484376,
|
| 444 |
+
"step": 1220
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.013142545613046097,
|
| 448 |
+
"grad_norm": 16.171775817871094,
|
| 449 |
+
"learning_rate": 4.9999054522615176e-05,
|
| 450 |
+
"loss": 28.09063720703125,
|
| 451 |
+
"step": 1240
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.013354522155192,
|
| 455 |
+
"grad_norm": 12.458906173706055,
|
| 456 |
+
"learning_rate": 4.999899572016877e-05,
|
| 457 |
+
"loss": 26.42186584472656,
|
| 458 |
+
"step": 1260
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.013566498697337906,
|
| 462 |
+
"grad_norm": 15.455846786499023,
|
| 463 |
+
"learning_rate": 4.999893514394113e-05,
|
| 464 |
+
"loss": 26.841815185546874,
|
| 465 |
+
"step": 1280
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.013778475239483811,
|
| 469 |
+
"grad_norm": 13.138788223266602,
|
| 470 |
+
"learning_rate": 4.9998872793936535e-05,
|
| 471 |
+
"loss": 25.769387817382814,
|
| 472 |
+
"step": 1300
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.013990451781629715,
|
| 476 |
+
"grad_norm": 11.507213592529297,
|
| 477 |
+
"learning_rate": 4.999880867015942e-05,
|
| 478 |
+
"loss": 25.382391357421874,
|
| 479 |
+
"step": 1320
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.01420242832377562,
|
| 483 |
+
"grad_norm": 12.166060447692871,
|
| 484 |
+
"learning_rate": 4.9998742772614334e-05,
|
| 485 |
+
"loss": 24.878656005859376,
|
| 486 |
+
"step": 1340
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.014414404865921524,
|
| 490 |
+
"grad_norm": 13.178064346313477,
|
| 491 |
+
"learning_rate": 4.999867510130596e-05,
|
| 492 |
+
"loss": 25.014637756347657,
|
| 493 |
+
"step": 1360
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.01462638140806743,
|
| 497 |
+
"grad_norm": 11.066803932189941,
|
| 498 |
+
"learning_rate": 4.999860565623908e-05,
|
| 499 |
+
"loss": 24.23482666015625,
|
| 500 |
+
"step": 1380
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.014838357950213335,
|
| 504 |
+
"grad_norm": 14.150485038757324,
|
| 505 |
+
"learning_rate": 4.999853443741864e-05,
|
| 506 |
+
"loss": 24.704800415039063,
|
| 507 |
+
"step": 1400
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.015050334492359239,
|
| 511 |
+
"grad_norm": 11.332118034362793,
|
| 512 |
+
"learning_rate": 4.999846144484969e-05,
|
| 513 |
+
"loss": 24.328228759765626,
|
| 514 |
+
"step": 1420
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.015262311034505144,
|
| 518 |
+
"grad_norm": 14.096946716308594,
|
| 519 |
+
"learning_rate": 4.99983866785374e-05,
|
| 520 |
+
"loss": 24.1265625,
|
| 521 |
+
"step": 1440
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.01547428757665105,
|
| 525 |
+
"grad_norm": 12.599404335021973,
|
| 526 |
+
"learning_rate": 4.999831013848709e-05,
|
| 527 |
+
"loss": 23.550830078125,
|
| 528 |
+
"step": 1460
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.015686264118796953,
|
| 532 |
+
"grad_norm": 10.074280738830566,
|
| 533 |
+
"learning_rate": 4.999823182470417e-05,
|
| 534 |
+
"loss": 22.190635681152344,
|
| 535 |
+
"step": 1480
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.015898240660942857,
|
| 539 |
+
"grad_norm": 11.788599014282227,
|
| 540 |
+
"learning_rate": 4.999815173719422e-05,
|
| 541 |
+
"loss": 22.380284118652344,
|
| 542 |
+
"step": 1500
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.016110217203088764,
|
| 546 |
+
"grad_norm": 11.708645820617676,
|
| 547 |
+
"learning_rate": 4.9998069875962904e-05,
|
| 548 |
+
"loss": 21.631423950195312,
|
| 549 |
+
"step": 1520
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.016322193745234668,
|
| 553 |
+
"grad_norm": 11.673741340637207,
|
| 554 |
+
"learning_rate": 4.9997986241016035e-05,
|
| 555 |
+
"loss": 21.9886474609375,
|
| 556 |
+
"step": 1540
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.01653417028738057,
|
| 560 |
+
"grad_norm": 11.298181533813477,
|
| 561 |
+
"learning_rate": 4.9997900832359556e-05,
|
| 562 |
+
"loss": 22.229173278808595,
|
| 563 |
+
"step": 1560
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.01674614682952648,
|
| 567 |
+
"grad_norm": 11.78747272491455,
|
| 568 |
+
"learning_rate": 4.999781364999952e-05,
|
| 569 |
+
"loss": 21.366239929199217,
|
| 570 |
+
"step": 1580
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.016958123371672382,
|
| 574 |
+
"grad_norm": 11.006595611572266,
|
| 575 |
+
"learning_rate": 4.9997724693942106e-05,
|
| 576 |
+
"loss": 20.945326232910155,
|
| 577 |
+
"step": 1600
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.017170099913818286,
|
| 581 |
+
"grad_norm": 17.829376220703125,
|
| 582 |
+
"learning_rate": 4.999763396419364e-05,
|
| 583 |
+
"loss": 21.32393798828125,
|
| 584 |
+
"step": 1620
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.017382076455964193,
|
| 588 |
+
"grad_norm": 10.26064395904541,
|
| 589 |
+
"learning_rate": 4.999754146076055e-05,
|
| 590 |
+
"loss": 20.603547668457033,
|
| 591 |
+
"step": 1640
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.017594052998110097,
|
| 595 |
+
"grad_norm": 9.404328346252441,
|
| 596 |
+
"learning_rate": 4.9997447183649406e-05,
|
| 597 |
+
"loss": 20.426235961914063,
|
| 598 |
+
"step": 1660
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.017806029540256,
|
| 602 |
+
"grad_norm": 13.227001190185547,
|
| 603 |
+
"learning_rate": 4.9997351132866885e-05,
|
| 604 |
+
"loss": 20.314698791503908,
|
| 605 |
+
"step": 1680
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.018018006082401908,
|
| 609 |
+
"grad_norm": 13.900830268859863,
|
| 610 |
+
"learning_rate": 4.999725330841982e-05,
|
| 611 |
+
"loss": 20.570407104492187,
|
| 612 |
+
"step": 1700
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.01822998262454781,
|
| 616 |
+
"grad_norm": 12.000020027160645,
|
| 617 |
+
"learning_rate": 4.999715371031515e-05,
|
| 618 |
+
"loss": 19.933245849609374,
|
| 619 |
+
"step": 1720
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.018441963238843298,
|
| 623 |
+
"grad_norm": 15.08840274810791,
|
| 624 |
+
"learning_rate": 4.999705233855993e-05,
|
| 625 |
+
"loss": 20.505574544270832,
|
| 626 |
+
"step": 1740
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.018653939827795517,
|
| 630 |
+
"grad_norm": 8.573169708251953,
|
| 631 |
+
"learning_rate": 4.999694919316135e-05,
|
| 632 |
+
"loss": 21.149876403808594,
|
| 633 |
+
"step": 1760
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.01886591641674774,
|
| 637 |
+
"grad_norm": 9.61663818359375,
|
| 638 |
+
"learning_rate": 4.999684427412674e-05,
|
| 639 |
+
"loss": 19.591824340820313,
|
| 640 |
+
"step": 1780
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.019077893005699963,
|
| 644 |
+
"grad_norm": 7.931303977966309,
|
| 645 |
+
"learning_rate": 4.999673758146354e-05,
|
| 646 |
+
"loss": 18.761630249023437,
|
| 647 |
+
"step": 1800
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.019289869594652183,
|
| 651 |
+
"grad_norm": 9.059905052185059,
|
| 652 |
+
"learning_rate": 4.999662911517932e-05,
|
| 653 |
+
"loss": 18.697032165527343,
|
| 654 |
+
"step": 1820
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.019501846183604406,
|
| 658 |
+
"grad_norm": 9.243865013122559,
|
| 659 |
+
"learning_rate": 4.999652442939985e-05,
|
| 660 |
+
"loss": 18.411224365234375,
|
| 661 |
+
"step": 1840
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.01971382277255663,
|
| 665 |
+
"grad_norm": 9.229769706726074,
|
| 666 |
+
"learning_rate": 4.99964125045769e-05,
|
| 667 |
+
"loss": 17.485566711425783,
|
| 668 |
+
"step": 1860
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.019925799361508848,
|
| 672 |
+
"grad_norm": 8.589388847351074,
|
| 673 |
+
"learning_rate": 4.999629880615599e-05,
|
| 674 |
+
"loss": 17.406051635742188,
|
| 675 |
+
"step": 1880
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.02013777595046107,
|
| 679 |
+
"grad_norm": 17.106897354125977,
|
| 680 |
+
"learning_rate": 4.9996183334145195e-05,
|
| 681 |
+
"loss": 16.668821716308592,
|
| 682 |
+
"step": 1900
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.020349752539413294,
|
| 686 |
+
"grad_norm": 8.545101165771484,
|
| 687 |
+
"learning_rate": 4.9996066088552715e-05,
|
| 688 |
+
"loss": 16.87046661376953,
|
| 689 |
+
"step": 1920
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.020561729128365513,
|
| 693 |
+
"grad_norm": 16.08205795288086,
|
| 694 |
+
"learning_rate": 4.999594706938685e-05,
|
| 695 |
+
"loss": 16.353001403808594,
|
| 696 |
+
"step": 1940
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.020773705717317736,
|
| 700 |
+
"grad_norm": 11.286654472351074,
|
| 701 |
+
"learning_rate": 4.999582627665606e-05,
|
| 702 |
+
"loss": 15.638079833984374,
|
| 703 |
+
"step": 1960
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.02098568230626996,
|
| 707 |
+
"grad_norm": 9.223414421081543,
|
| 708 |
+
"learning_rate": 4.999570371036891e-05,
|
| 709 |
+
"loss": 15.506455993652343,
|
| 710 |
+
"step": 1980
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.02119765889522218,
|
| 714 |
+
"grad_norm": 8.263184547424316,
|
| 715 |
+
"learning_rate": 4.9995579370534104e-05,
|
| 716 |
+
"loss": 14.92587890625,
|
| 717 |
+
"step": 2000
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.021409635484174402,
|
| 721 |
+
"grad_norm": 14.854691505432129,
|
| 722 |
+
"learning_rate": 4.999545325716046e-05,
|
| 723 |
+
"loss": 15.989543151855468,
|
| 724 |
+
"step": 2020
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 0.021621612073126625,
|
| 728 |
+
"grad_norm": 8.34041976928711,
|
| 729 |
+
"learning_rate": 4.999532537025692e-05,
|
| 730 |
+
"loss": 14.783149719238281,
|
| 731 |
+
"step": 2040
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.021833588662078844,
|
| 735 |
+
"grad_norm": 7.618574619293213,
|
| 736 |
+
"learning_rate": 4.999519570983257e-05,
|
| 737 |
+
"loss": 14.854769897460937,
|
| 738 |
+
"step": 2060
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.022045565251031067,
|
| 742 |
+
"grad_norm": 8.25412368774414,
|
| 743 |
+
"learning_rate": 4.9995064275896595e-05,
|
| 744 |
+
"loss": 14.659225463867188,
|
| 745 |
+
"step": 2080
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.02225754183998329,
|
| 749 |
+
"grad_norm": 8.886482238769531,
|
| 750 |
+
"learning_rate": 4.999493106845833e-05,
|
| 751 |
+
"loss": 14.271562194824218,
|
| 752 |
+
"step": 2100
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.02246951842893551,
|
| 756 |
+
"grad_norm": 18.914173126220703,
|
| 757 |
+
"learning_rate": 4.999479608752723e-05,
|
| 758 |
+
"loss": 14.569769287109375,
|
| 759 |
+
"step": 2120
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.022681495017887732,
|
| 763 |
+
"grad_norm": 9.543008804321289,
|
| 764 |
+
"learning_rate": 4.999465933311286e-05,
|
| 765 |
+
"loss": 14.546726989746094,
|
| 766 |
+
"step": 2140
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 0.022893471606839955,
|
| 770 |
+
"grad_norm": 15.414135932922363,
|
| 771 |
+
"learning_rate": 4.9994520805224944e-05,
|
| 772 |
+
"loss": 14.693588256835938,
|
| 773 |
+
"step": 2160
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.023105448195792175,
|
| 777 |
+
"grad_norm": 7.707723140716553,
|
| 778 |
+
"learning_rate": 4.999438050387329e-05,
|
| 779 |
+
"loss": 13.439755249023438,
|
| 780 |
+
"step": 2180
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 0.023317424784744398,
|
| 784 |
+
"grad_norm": 8.205167770385742,
|
| 785 |
+
"learning_rate": 4.999423842906786e-05,
|
| 786 |
+
"loss": 13.753311157226562,
|
| 787 |
+
"step": 2200
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 0.02352940137369662,
|
| 791 |
+
"grad_norm": 8.64995002746582,
|
| 792 |
+
"learning_rate": 4.999409458081874e-05,
|
| 793 |
+
"loss": 13.15582733154297,
|
| 794 |
+
"step": 2220
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.02374137796264884,
|
| 798 |
+
"grad_norm": 10.525213241577148,
|
| 799 |
+
"learning_rate": 4.999394895913613e-05,
|
| 800 |
+
"loss": 13.306376647949218,
|
| 801 |
+
"step": 2240
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.023953354551601063,
|
| 805 |
+
"grad_norm": 8.105880737304688,
|
| 806 |
+
"learning_rate": 4.9993801564030375e-05,
|
| 807 |
+
"loss": 13.052978515625,
|
| 808 |
+
"step": 2260
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 0.024165331140553286,
|
| 812 |
+
"grad_norm": 9.805683135986328,
|
| 813 |
+
"learning_rate": 4.9993652395511915e-05,
|
| 814 |
+
"loss": 13.548892211914062,
|
| 815 |
+
"step": 2280
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 0.02437730772950551,
|
| 819 |
+
"grad_norm": 7.142966270446777,
|
| 820 |
+
"learning_rate": 4.999350145359134e-05,
|
| 821 |
+
"loss": 13.522245788574219,
|
| 822 |
+
"step": 2300
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 0.02458928431845773,
|
| 826 |
+
"grad_norm": 9.679190635681152,
|
| 827 |
+
"learning_rate": 4.9993348738279364e-05,
|
| 828 |
+
"loss": 13.134727478027344,
|
| 829 |
+
"step": 2320
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 0.02480126090740995,
|
| 833 |
+
"grad_norm": 7.862600803375244,
|
| 834 |
+
"learning_rate": 4.999319424958683e-05,
|
| 835 |
+
"loss": 13.142591857910157,
|
| 836 |
+
"step": 2340
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.025013237496362174,
|
| 840 |
+
"grad_norm": 9.43050765991211,
|
| 841 |
+
"learning_rate": 4.999304584274515e-05,
|
| 842 |
+
"loss": 13.02917022705078,
|
| 843 |
+
"step": 2360
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 0.025225214085314394,
|
| 847 |
+
"grad_norm": 10.131393432617188,
|
| 848 |
+
"learning_rate": 4.999288789599215e-05,
|
| 849 |
+
"loss": 12.05879898071289,
|
| 850 |
+
"step": 2380
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 0.025437190674266617,
|
| 854 |
+
"grad_norm": 9.216246604919434,
|
| 855 |
+
"learning_rate": 4.999272817589127e-05,
|
| 856 |
+
"loss": 12.294894409179687,
|
| 857 |
+
"step": 2400
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 0.02564916726321884,
|
| 861 |
+
"grad_norm": 7.373030662536621,
|
| 862 |
+
"learning_rate": 4.999256668245387e-05,
|
| 863 |
+
"loss": 12.136451721191406,
|
| 864 |
+
"step": 2420
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 0.02586114385217106,
|
| 868 |
+
"grad_norm": 7.433055877685547,
|
| 869 |
+
"learning_rate": 4.999240341569139e-05,
|
| 870 |
+
"loss": 11.948836517333984,
|
| 871 |
+
"step": 2440
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 0.026073120441123282,
|
| 875 |
+
"grad_norm": 7.93651008605957,
|
| 876 |
+
"learning_rate": 4.999223837561541e-05,
|
| 877 |
+
"loss": 12.189994812011719,
|
| 878 |
+
"step": 2460
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 0.026285097030075505,
|
| 882 |
+
"grad_norm": 9.212141036987305,
|
| 883 |
+
"learning_rate": 4.9992071562237654e-05,
|
| 884 |
+
"loss": 11.8706787109375,
|
| 885 |
+
"step": 2480
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 0.026497073619027724,
|
| 889 |
+
"grad_norm": 9.067305564880371,
|
| 890 |
+
"learning_rate": 4.999190297556995e-05,
|
| 891 |
+
"loss": 11.95904769897461,
|
| 892 |
+
"step": 2500
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 0.026709050207979947,
|
| 896 |
+
"grad_norm": 6.703977584838867,
|
| 897 |
+
"learning_rate": 4.999173261562426e-05,
|
| 898 |
+
"loss": 11.5086181640625,
|
| 899 |
+
"step": 2520
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 0.026921020852529908,
|
| 903 |
+
"grad_norm": 14.173149108886719,
|
| 904 |
+
"learning_rate": 4.999156048241269e-05,
|
| 905 |
+
"loss": 14.55417960030692,
|
| 906 |
+
"step": 2540
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 0.027132997394675812,
|
| 910 |
+
"grad_norm": 9.45891284942627,
|
| 911 |
+
"learning_rate": 4.999138657594742e-05,
|
| 912 |
+
"loss": 13.479953002929687,
|
| 913 |
+
"step": 2560
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 0.027344973936821716,
|
| 917 |
+
"grad_norm": 8.567397117614746,
|
| 918 |
+
"learning_rate": 4.9991210896240815e-05,
|
| 919 |
+
"loss": 12.890872192382812,
|
| 920 |
+
"step": 2580
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 0.027556950478967623,
|
| 924 |
+
"grad_norm": 9.527186393737793,
|
| 925 |
+
"learning_rate": 4.999103344330533e-05,
|
| 926 |
+
"loss": 12.856292724609375,
|
| 927 |
+
"step": 2600
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 0.027768927021113526,
|
| 931 |
+
"grad_norm": 10.341022491455078,
|
| 932 |
+
"learning_rate": 4.999085421715356e-05,
|
| 933 |
+
"loss": 12.83624267578125,
|
| 934 |
+
"step": 2620
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 0.02798090356325943,
|
| 938 |
+
"grad_norm": 9.412083625793457,
|
| 939 |
+
"learning_rate": 4.9990673217798214e-05,
|
| 940 |
+
"loss": 13.19751434326172,
|
| 941 |
+
"step": 2640
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 0.028192880105405337,
|
| 945 |
+
"grad_norm": 7.499392986297607,
|
| 946 |
+
"learning_rate": 4.999049044525215e-05,
|
| 947 |
+
"loss": 12.54099349975586,
|
| 948 |
+
"step": 2660
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 0.02840485664755124,
|
| 952 |
+
"grad_norm": 8.639656066894531,
|
| 953 |
+
"learning_rate": 4.999030589952832e-05,
|
| 954 |
+
"loss": 12.512214660644531,
|
| 955 |
+
"step": 2680
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 0.028616833189697145,
|
| 959 |
+
"grad_norm": 7.731313228607178,
|
| 960 |
+
"learning_rate": 4.9990119580639825e-05,
|
| 961 |
+
"loss": 12.74936294555664,
|
| 962 |
+
"step": 2700
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 0.02882880973184305,
|
| 966 |
+
"grad_norm": 8.4285888671875,
|
| 967 |
+
"learning_rate": 4.998993148859988e-05,
|
| 968 |
+
"loss": 12.302136993408203,
|
| 969 |
+
"step": 2720
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 0.029040786273988956,
|
| 973 |
+
"grad_norm": 8.15134334564209,
|
| 974 |
+
"learning_rate": 4.9989741623421845e-05,
|
| 975 |
+
"loss": 12.26800765991211,
|
| 976 |
+
"step": 2740
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 0.02925276281613486,
|
| 980 |
+
"grad_norm": 10.479498863220215,
|
| 981 |
+
"learning_rate": 4.9989549985119186e-05,
|
| 982 |
+
"loss": 11.683930969238281,
|
| 983 |
+
"step": 2760
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 0.029464739358280763,
|
| 987 |
+
"grad_norm": 9.25798225402832,
|
| 988 |
+
"learning_rate": 4.998935657370549e-05,
|
| 989 |
+
"loss": 12.325420379638672,
|
| 990 |
+
"step": 2780
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 0.02967671590042667,
|
| 994 |
+
"grad_norm": 10.919699668884277,
|
| 995 |
+
"learning_rate": 4.9989161389194494e-05,
|
| 996 |
+
"loss": 12.542608642578125,
|
| 997 |
+
"step": 2800
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 0.029888692442572574,
|
| 1001 |
+
"grad_norm": 7.847535610198975,
|
| 1002 |
+
"learning_rate": 4.998896443160004e-05,
|
| 1003 |
+
"loss": 11.772618103027344,
|
| 1004 |
+
"step": 2820
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 0.030100668984718477,
|
| 1008 |
+
"grad_norm": 6.975454807281494,
|
| 1009 |
+
"learning_rate": 4.99887657009361e-05,
|
| 1010 |
+
"loss": 11.24947509765625,
|
| 1011 |
+
"step": 2840
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 0.030312645526864385,
|
| 1015 |
+
"grad_norm": 11.4141845703125,
|
| 1016 |
+
"learning_rate": 4.998856519721678e-05,
|
| 1017 |
+
"loss": 11.529437255859374,
|
| 1018 |
+
"step": 2860
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 0.03052462206901029,
|
| 1022 |
+
"grad_norm": 8.164746284484863,
|
| 1023 |
+
"learning_rate": 4.99883629204563e-05,
|
| 1024 |
+
"loss": 11.559315490722657,
|
| 1025 |
+
"step": 2880
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 0.030736598611156192,
|
| 1029 |
+
"grad_norm": 10.127477645874023,
|
| 1030 |
+
"learning_rate": 4.9988158870669024e-05,
|
| 1031 |
+
"loss": 11.085247039794922,
|
| 1032 |
+
"step": 2900
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 0.0309485751533021,
|
| 1036 |
+
"grad_norm": 7.732602596282959,
|
| 1037 |
+
"learning_rate": 4.9987953047869425e-05,
|
| 1038 |
+
"loss": 11.550102996826173,
|
| 1039 |
+
"step": 2920
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 0.031160551695448003,
|
| 1043 |
+
"grad_norm": 9.073702812194824,
|
| 1044 |
+
"learning_rate": 4.99877454520721e-05,
|
| 1045 |
+
"loss": 10.96413345336914,
|
| 1046 |
+
"step": 2940
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 0.03137252823759391,
|
| 1050 |
+
"grad_norm": 7.6227521896362305,
|
| 1051 |
+
"learning_rate": 4.998753608329179e-05,
|
| 1052 |
+
"loss": 10.918690490722657,
|
| 1053 |
+
"step": 2960
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 0.031584504779739814,
|
| 1057 |
+
"grad_norm": 8.286331176757812,
|
| 1058 |
+
"learning_rate": 4.9987324941543346e-05,
|
| 1059 |
+
"loss": 11.397634887695313,
|
| 1060 |
+
"step": 2980
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 0.031796481321885714,
|
| 1064 |
+
"grad_norm": 7.026095390319824,
|
| 1065 |
+
"learning_rate": 4.998711202684174e-05,
|
| 1066 |
+
"loss": 11.12390594482422,
|
| 1067 |
+
"step": 3000
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 0.03200845786403162,
|
| 1071 |
+
"grad_norm": 8.56346321105957,
|
| 1072 |
+
"learning_rate": 4.99868973392021e-05,
|
| 1073 |
+
"loss": 10.637153625488281,
|
| 1074 |
+
"step": 3020
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 0.03222043440617753,
|
| 1078 |
+
"grad_norm": 8.143285751342773,
|
| 1079 |
+
"learning_rate": 4.9986680878639635e-05,
|
| 1080 |
+
"loss": 10.714803314208984,
|
| 1081 |
+
"step": 3040
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 0.03243241094832343,
|
| 1085 |
+
"grad_norm": 7.185257434844971,
|
| 1086 |
+
"learning_rate": 4.998646264516972e-05,
|
| 1087 |
+
"loss": 10.133228302001953,
|
| 1088 |
+
"step": 3060
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 0.032644387490469336,
|
| 1092 |
+
"grad_norm": 7.972578525543213,
|
| 1093 |
+
"learning_rate": 4.9986242638807835e-05,
|
| 1094 |
+
"loss": 10.702585601806641,
|
| 1095 |
+
"step": 3080
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 0.03285636403261524,
|
| 1099 |
+
"grad_norm": 7.698553085327148,
|
| 1100 |
+
"learning_rate": 4.998602085956959e-05,
|
| 1101 |
+
"loss": 10.520194244384765,
|
| 1102 |
+
"step": 3100
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 0.03306834057476114,
|
| 1106 |
+
"grad_norm": 8.99880599975586,
|
| 1107 |
+
"learning_rate": 4.998579730747073e-05,
|
| 1108 |
+
"loss": 10.247818756103516,
|
| 1109 |
+
"step": 3120
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 0.03328031711690705,
|
| 1113 |
+
"grad_norm": 11.4783296585083,
|
| 1114 |
+
"learning_rate": 4.9985571982527094e-05,
|
| 1115 |
+
"loss": 10.400140380859375,
|
| 1116 |
+
"step": 3140
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 0.03349229365905296,
|
| 1120 |
+
"grad_norm": 7.925187110900879,
|
| 1121 |
+
"learning_rate": 4.998534488475469e-05,
|
| 1122 |
+
"loss": 10.009452056884765,
|
| 1123 |
+
"step": 3160
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 0.03370427020119886,
|
| 1127 |
+
"grad_norm": 16.27712059020996,
|
| 1128 |
+
"learning_rate": 4.998511601416963e-05,
|
| 1129 |
+
"loss": 10.09621810913086,
|
| 1130 |
+
"step": 3180
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 0.033916246743344765,
|
| 1134 |
+
"grad_norm": 8.849685668945312,
|
| 1135 |
+
"learning_rate": 4.998488537078814e-05,
|
| 1136 |
+
"loss": 10.090743255615234,
|
| 1137 |
+
"step": 3200
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 0.03412822328549067,
|
| 1141 |
+
"grad_norm": 8.597989082336426,
|
| 1142 |
+
"learning_rate": 4.99846529546266e-05,
|
| 1143 |
+
"loss": 10.238027954101563,
|
| 1144 |
+
"step": 3220
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 0.03434019982763657,
|
| 1148 |
+
"grad_norm": 9.006299018859863,
|
| 1149 |
+
"learning_rate": 4.99844187657015e-05,
|
| 1150 |
+
"loss": 10.442254638671875,
|
| 1151 |
+
"step": 3240
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 0.03455217636978248,
|
| 1155 |
+
"grad_norm": 7.547255992889404,
|
| 1156 |
+
"learning_rate": 4.998418280402944e-05,
|
| 1157 |
+
"loss": 10.080032348632812,
|
| 1158 |
+
"step": 3260
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 0.034764152911928387,
|
| 1162 |
+
"grad_norm": 8.353964805603027,
|
| 1163 |
+
"learning_rate": 4.998394506962718e-05,
|
| 1164 |
+
"loss": 9.597069549560548,
|
| 1165 |
+
"step": 3280
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 0.03497612945407429,
|
| 1169 |
+
"grad_norm": 6.397212028503418,
|
| 1170 |
+
"learning_rate": 4.9983705562511586e-05,
|
| 1171 |
+
"loss": 10.105459594726563,
|
| 1172 |
+
"step": 3300
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 0.035188105996220194,
|
| 1176 |
+
"grad_norm": 7.432406902313232,
|
| 1177 |
+
"learning_rate": 4.9983464282699647e-05,
|
| 1178 |
+
"loss": 9.653330993652343,
|
| 1179 |
+
"step": 3320
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 0.0354000825383661,
|
| 1183 |
+
"grad_norm": 7.640140056610107,
|
| 1184 |
+
"learning_rate": 4.9983221230208475e-05,
|
| 1185 |
+
"loss": 9.88272476196289,
|
| 1186 |
+
"step": 3340
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 0.035612059080512,
|
| 1190 |
+
"grad_norm": 7.367326259613037,
|
| 1191 |
+
"learning_rate": 4.998297640505533e-05,
|
| 1192 |
+
"loss": 9.75995101928711,
|
| 1193 |
+
"step": 3360
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 0.03582403562265791,
|
| 1197 |
+
"grad_norm": 7.907365798950195,
|
| 1198 |
+
"learning_rate": 4.998272980725758e-05,
|
| 1199 |
+
"loss": 10.62704086303711,
|
| 1200 |
+
"step": 3380
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 0.036036012164803816,
|
| 1204 |
+
"grad_norm": 6.930318355560303,
|
| 1205 |
+
"learning_rate": 4.998248143683272e-05,
|
| 1206 |
+
"loss": 9.59298095703125,
|
| 1207 |
+
"step": 3400
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 0.036247988706949716,
|
| 1211 |
+
"grad_norm": 6.814286231994629,
|
| 1212 |
+
"learning_rate": 4.998223129379837e-05,
|
| 1213 |
+
"loss": 9.595169067382812,
|
| 1214 |
+
"step": 3420
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 0.03645996524909562,
|
| 1218 |
+
"grad_norm": 11.377059936523438,
|
| 1219 |
+
"learning_rate": 4.9981979378172285e-05,
|
| 1220 |
+
"loss": 9.344342803955078,
|
| 1221 |
+
"step": 3440
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 0.03667194179124153,
|
| 1225 |
+
"grad_norm": 8.139765739440918,
|
| 1226 |
+
"learning_rate": 4.9981725689972335e-05,
|
| 1227 |
+
"loss": 9.5405029296875,
|
| 1228 |
+
"step": 3460
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 0.03688391833338743,
|
| 1232 |
+
"grad_norm": 6.915637016296387,
|
| 1233 |
+
"learning_rate": 4.998147022921652e-05,
|
| 1234 |
+
"loss": 9.035648345947266,
|
| 1235 |
+
"step": 3480
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 0.03709589487553334,
|
| 1239 |
+
"grad_norm": 10.19848346710205,
|
| 1240 |
+
"learning_rate": 4.9981212995922963e-05,
|
| 1241 |
+
"loss": 9.29233627319336,
|
| 1242 |
+
"step": 3500
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 0.03730787141767924,
|
| 1246 |
+
"grad_norm": 8.867091178894043,
|
| 1247 |
+
"learning_rate": 4.998095399010993e-05,
|
| 1248 |
+
"loss": 9.588215637207032,
|
| 1249 |
+
"step": 3520
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 0.037519847959825145,
|
| 1253 |
+
"grad_norm": 9.17452621459961,
|
| 1254 |
+
"learning_rate": 4.998069321179578e-05,
|
| 1255 |
+
"loss": 9.314067840576172,
|
| 1256 |
+
"step": 3540
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 0.03773182450197105,
|
| 1260 |
+
"grad_norm": 7.039797306060791,
|
| 1261 |
+
"learning_rate": 4.9980430660999035e-05,
|
| 1262 |
+
"loss": 8.808745574951171,
|
| 1263 |
+
"step": 3560
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 0.03794380104411695,
|
| 1267 |
+
"grad_norm": 13.770051956176758,
|
| 1268 |
+
"learning_rate": 4.998016633773831e-05,
|
| 1269 |
+
"loss": 8.580646514892578,
|
| 1270 |
+
"step": 3580
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 0.03815577758626286,
|
| 1274 |
+
"grad_norm": 9.368520736694336,
|
| 1275 |
+
"learning_rate": 4.997990024203236e-05,
|
| 1276 |
+
"loss": 8.874877166748046,
|
| 1277 |
+
"step": 3600
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 0.03836775412840877,
|
| 1281 |
+
"grad_norm": 6.9515509605407715,
|
| 1282 |
+
"learning_rate": 4.997963237390007e-05,
|
| 1283 |
+
"loss": 8.784373474121093,
|
| 1284 |
+
"step": 3620
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 0.03857973067055467,
|
| 1288 |
+
"grad_norm": 17.312786102294922,
|
| 1289 |
+
"learning_rate": 4.997936273336045e-05,
|
| 1290 |
+
"loss": 8.654251861572266,
|
| 1291 |
+
"step": 3640
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 0.038791707212700574,
|
| 1295 |
+
"grad_norm": 7.585219860076904,
|
| 1296 |
+
"learning_rate": 4.997909132043262e-05,
|
| 1297 |
+
"loss": 8.881016540527344,
|
| 1298 |
+
"step": 3660
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 0.03900368375484648,
|
| 1302 |
+
"grad_norm": 6.534177780151367,
|
| 1303 |
+
"learning_rate": 4.997881813513585e-05,
|
| 1304 |
+
"loss": 8.644546508789062,
|
| 1305 |
+
"step": 3680
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 0.03921566029699238,
|
| 1309 |
+
"grad_norm": 7.262326717376709,
|
| 1310 |
+
"learning_rate": 4.997854317748952e-05,
|
| 1311 |
+
"loss": 8.438851165771485,
|
| 1312 |
+
"step": 3700
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 0.03942763683913829,
|
| 1316 |
+
"grad_norm": 6.688304901123047,
|
| 1317 |
+
"learning_rate": 4.9978266447513136e-05,
|
| 1318 |
+
"loss": 8.818517303466797,
|
| 1319 |
+
"step": 3720
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 0.039639613381284196,
|
| 1323 |
+
"grad_norm": 7.96449613571167,
|
| 1324 |
+
"learning_rate": 4.997798794522633e-05,
|
| 1325 |
+
"loss": 8.193695831298829,
|
| 1326 |
+
"step": 3740
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 0.039851589923430096,
|
| 1330 |
+
"grad_norm": 7.225415229797363,
|
| 1331 |
+
"learning_rate": 4.997770767064889e-05,
|
| 1332 |
+
"loss": 8.719413757324219,
|
| 1333 |
+
"step": 3760
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 0.040063566465576,
|
| 1337 |
+
"grad_norm": 10.660993576049805,
|
| 1338 |
+
"learning_rate": 4.997742562380067e-05,
|
| 1339 |
+
"loss": 8.99606704711914,
|
| 1340 |
+
"step": 3780
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 0.04027554300772191,
|
| 1344 |
+
"grad_norm": 7.489035129547119,
|
| 1345 |
+
"learning_rate": 4.997714180470168e-05,
|
| 1346 |
+
"loss": 8.775997924804688,
|
| 1347 |
+
"step": 3800
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 0.04048751954986781,
|
| 1351 |
+
"grad_norm": 7.120395183563232,
|
| 1352 |
+
"learning_rate": 4.9976856213372084e-05,
|
| 1353 |
+
"loss": 8.550077819824219,
|
| 1354 |
+
"step": 3820
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 0.04069949609201372,
|
| 1358 |
+
"grad_norm": 6.620581150054932,
|
| 1359 |
+
"learning_rate": 4.997656884983214e-05,
|
| 1360 |
+
"loss": 8.073123168945312,
|
| 1361 |
+
"step": 3840
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 0.040911472634159625,
|
| 1365 |
+
"grad_norm": 6.888649940490723,
|
| 1366 |
+
"learning_rate": 4.997627971410222e-05,
|
| 1367 |
+
"loss": 8.246237182617188,
|
| 1368 |
+
"step": 3860
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 0.03426954098025461,
|
| 1372 |
+
"grad_norm": 67.43865203857422,
|
| 1373 |
+
"learning_rate": 5e-05,
|
| 1374 |
+
"loss": 5.823767185211182,
|
| 1375 |
+
"step": 3880
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 0.03444618809870953,
|
| 1379 |
+
"grad_norm": 197.56736755371094,
|
| 1380 |
+
"learning_rate": 4.9999999846015945e-05,
|
| 1381 |
+
"loss": 26.320828247070313,
|
| 1382 |
+
"step": 3900
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 0.04154741143463547,
|
| 1386 |
+
"grad_norm": 9.097684860229492,
|
| 1387 |
+
"learning_rate": 4.999999994456711e-05,
|
| 1388 |
+
"loss": 11.287847137451172,
|
| 1389 |
+
"step": 3920
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 0.041759388023587696,
|
| 1393 |
+
"grad_norm": 7.985315322875977,
|
| 1394 |
+
"learning_rate": 4.999999861417774e-05,
|
| 1395 |
+
"loss": 10.503189086914062,
|
| 1396 |
+
"step": 3940
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 0.04197136461253992,
|
| 1400 |
+
"grad_norm": 7.371786117553711,
|
| 1401 |
+
"learning_rate": 4.999999550993597e-05,
|
| 1402 |
+
"loss": 10.253286743164063,
|
| 1403 |
+
"step": 3960
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 0.04218334120149214,
|
| 1407 |
+
"grad_norm": 8.781425476074219,
|
| 1408 |
+
"learning_rate": 4.9999990631842006e-05,
|
| 1409 |
+
"loss": 9.602980804443359,
|
| 1410 |
+
"step": 3980
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 0.04239531779044436,
|
| 1414 |
+
"grad_norm": 7.48057746887207,
|
| 1415 |
+
"learning_rate": 4.999998397989622e-05,
|
| 1416 |
+
"loss": 9.229662322998047,
|
| 1417 |
+
"step": 4000
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 0.04260729437939658,
|
| 1421 |
+
"grad_norm": 8.220187187194824,
|
| 1422 |
+
"learning_rate": 4.9999975554099065e-05,
|
| 1423 |
+
"loss": 9.642359924316406,
|
| 1424 |
+
"step": 4020
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 0.042819270968348803,
|
| 1428 |
+
"grad_norm": 7.194672107696533,
|
| 1429 |
+
"learning_rate": 4.999996535445114e-05,
|
| 1430 |
+
"loss": 9.438529968261719,
|
| 1431 |
+
"step": 4040
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 0.043031247557301026,
|
| 1435 |
+
"grad_norm": 7.428903579711914,
|
| 1436 |
+
"learning_rate": 4.999995338095318e-05,
|
| 1437 |
+
"loss": 9.112210845947265,
|
| 1438 |
+
"step": 4060
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 0.04324322414625325,
|
| 1442 |
+
"grad_norm": 7.56043815612793,
|
| 1443 |
+
"learning_rate": 4.999993963360602e-05,
|
| 1444 |
+
"loss": 8.96917266845703,
|
| 1445 |
+
"step": 4080
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 0.04345520073520547,
|
| 1449 |
+
"grad_norm": 7.234394550323486,
|
| 1450 |
+
"learning_rate": 4.9999924112410645e-05,
|
| 1451 |
+
"loss": 8.692044830322265,
|
| 1452 |
+
"step": 4100
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 0.04366717732415769,
|
| 1456 |
+
"grad_norm": 8.71029281616211,
|
| 1457 |
+
"learning_rate": 4.9999906817368144e-05,
|
| 1458 |
+
"loss": 9.252363586425782,
|
| 1459 |
+
"step": 4120
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 0.04387915391310991,
|
| 1463 |
+
"grad_norm": 7.397716045379639,
|
| 1464 |
+
"learning_rate": 4.999988774847977e-05,
|
| 1465 |
+
"loss": 8.999927520751953,
|
| 1466 |
+
"step": 4140
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 0.044091130502062134,
|
| 1470 |
+
"grad_norm": 7.631270885467529,
|
| 1471 |
+
"learning_rate": 4.999986690574685e-05,
|
| 1472 |
+
"loss": 8.953203582763672,
|
| 1473 |
+
"step": 4160
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 0.04430310709101436,
|
| 1477 |
+
"grad_norm": 7.429250717163086,
|
| 1478 |
+
"learning_rate": 4.999984428917088e-05,
|
| 1479 |
+
"loss": 8.433536529541016,
|
| 1480 |
+
"step": 4180
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 0.04451508367996658,
|
| 1484 |
+
"grad_norm": 7.666223049163818,
|
| 1485 |
+
"learning_rate": 4.999981989875345e-05,
|
| 1486 |
+
"loss": 8.639158630371094,
|
| 1487 |
+
"step": 4200
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 0.0447270602689188,
|
| 1491 |
+
"grad_norm": 7.98890495300293,
|
| 1492 |
+
"learning_rate": 4.99997937344963e-05,
|
| 1493 |
+
"loss": 8.765853881835938,
|
| 1494 |
+
"step": 4220
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 0.04493903685787102,
|
| 1498 |
+
"grad_norm": 9.630439758300781,
|
| 1499 |
+
"learning_rate": 4.999976579640129e-05,
|
| 1500 |
+
"loss": 9.029901123046875,
|
| 1501 |
+
"step": 4240
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 0.04515101344682324,
|
| 1505 |
+
"grad_norm": 8.23205280303955,
|
| 1506 |
+
"learning_rate": 4.999973608447039e-05,
|
| 1507 |
+
"loss": 8.24354248046875,
|
| 1508 |
+
"step": 4260
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 0.045362990035775465,
|
| 1512 |
+
"grad_norm": 7.490217685699463,
|
| 1513 |
+
"learning_rate": 4.9999704598705716e-05,
|
| 1514 |
+
"loss": 8.166352081298829,
|
| 1515 |
+
"step": 4280
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 0.04557496662472769,
|
| 1519 |
+
"grad_norm": 6.790397644042969,
|
| 1520 |
+
"learning_rate": 4.99996713391095e-05,
|
| 1521 |
+
"loss": 7.992810821533203,
|
| 1522 |
+
"step": 4300
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 0.04578694321367991,
|
| 1526 |
+
"grad_norm": 6.565227031707764,
|
| 1527 |
+
"learning_rate": 4.999963630568411e-05,
|
| 1528 |
+
"loss": 7.932354736328125,
|
| 1529 |
+
"step": 4320
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 0.045998919802632134,
|
| 1533 |
+
"grad_norm": 7.678412437438965,
|
| 1534 |
+
"learning_rate": 4.999959949843202e-05,
|
| 1535 |
+
"loss": 8.180339813232422,
|
| 1536 |
+
"step": 4340
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 0.04621089639158435,
|
| 1540 |
+
"grad_norm": 8.342945098876953,
|
| 1541 |
+
"learning_rate": 4.9999560917355844e-05,
|
| 1542 |
+
"loss": 7.889363098144531,
|
| 1543 |
+
"step": 4360
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 0.04642287298053657,
|
| 1547 |
+
"grad_norm": 7.9554643630981445,
|
| 1548 |
+
"learning_rate": 4.9999520562458326e-05,
|
| 1549 |
+
"loss": 7.923197937011719,
|
| 1550 |
+
"step": 4380
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 0.046634849569488795,
|
| 1554 |
+
"grad_norm": 6.874939441680908,
|
| 1555 |
+
"learning_rate": 4.999947843374232e-05,
|
| 1556 |
+
"loss": 7.8224853515625,
|
| 1557 |
+
"step": 4400
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 0.04684682615844102,
|
| 1561 |
+
"grad_norm": 6.616832733154297,
|
| 1562 |
+
"learning_rate": 4.999943453121082e-05,
|
| 1563 |
+
"loss": 8.1573974609375,
|
| 1564 |
+
"step": 4420
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 0.04705880274739324,
|
| 1568 |
+
"grad_norm": 7.236156463623047,
|
| 1569 |
+
"learning_rate": 4.9999388854866947e-05,
|
| 1570 |
+
"loss": 7.864684295654297,
|
| 1571 |
+
"step": 4440
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 0.047270779336345464,
|
| 1575 |
+
"grad_norm": 7.412804126739502,
|
| 1576 |
+
"learning_rate": 4.9999341404713934e-05,
|
| 1577 |
+
"loss": 7.727866363525391,
|
| 1578 |
+
"step": 4460
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 0.04748275592529768,
|
| 1582 |
+
"grad_norm": 8.068159103393555,
|
| 1583 |
+
"learning_rate": 4.9999292180755156e-05,
|
| 1584 |
+
"loss": 7.871466827392578,
|
| 1585 |
+
"step": 4480
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 0.0476947325142499,
|
| 1589 |
+
"grad_norm": 6.552003860473633,
|
| 1590 |
+
"learning_rate": 4.999924118299409e-05,
|
| 1591 |
+
"loss": 7.4598548889160154,
|
| 1592 |
+
"step": 4500
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 0.047906709103202126,
|
| 1596 |
+
"grad_norm": 7.805352210998535,
|
| 1597 |
+
"learning_rate": 4.999918841143437e-05,
|
| 1598 |
+
"loss": 7.6653083801269535,
|
| 1599 |
+
"step": 4520
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 0.04811868569215435,
|
| 1603 |
+
"grad_norm": 7.179927825927734,
|
| 1604 |
+
"learning_rate": 4.999913386607973e-05,
|
| 1605 |
+
"loss": 7.798252868652344,
|
| 1606 |
+
"step": 4540
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 0.04833066228110657,
|
| 1610 |
+
"grad_norm": 7.078813076019287,
|
| 1611 |
+
"learning_rate": 4.999907754693405e-05,
|
| 1612 |
+
"loss": 7.617889404296875,
|
| 1613 |
+
"step": 4560
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 0.048542638870058795,
|
| 1617 |
+
"grad_norm": 7.527862071990967,
|
| 1618 |
+
"learning_rate": 4.999901945400132e-05,
|
| 1619 |
+
"loss": 7.342926025390625,
|
| 1620 |
+
"step": 4580
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 0.04875461545901102,
|
| 1624 |
+
"grad_norm": 6.760367393493652,
|
| 1625 |
+
"learning_rate": 4.999895958728566e-05,
|
| 1626 |
+
"loss": 7.390692901611328,
|
| 1627 |
+
"step": 4600
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 0.048966592047963234,
|
| 1631 |
+
"grad_norm": 7.048969745635986,
|
| 1632 |
+
"learning_rate": 4.999889794679132e-05,
|
| 1633 |
+
"loss": 7.3703147888183596,
|
| 1634 |
+
"step": 4620
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 0.04917856863691546,
|
| 1638 |
+
"grad_norm": 6.764970302581787,
|
| 1639 |
+
"learning_rate": 4.9998834532522674e-05,
|
| 1640 |
+
"loss": 7.666373443603516,
|
| 1641 |
+
"step": 4640
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 0.04939054522586768,
|
| 1645 |
+
"grad_norm": 7.154876708984375,
|
| 1646 |
+
"learning_rate": 4.999876934448423e-05,
|
| 1647 |
+
"loss": 7.437535095214844,
|
| 1648 |
+
"step": 4660
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 0.0496025218148199,
|
| 1652 |
+
"grad_norm": 7.958277702331543,
|
| 1653 |
+
"learning_rate": 4.9998702382680593e-05,
|
| 1654 |
+
"loss": 7.395094299316407,
|
| 1655 |
+
"step": 4680
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 0.049814498403772126,
|
| 1659 |
+
"grad_norm": 6.902902126312256,
|
| 1660 |
+
"learning_rate": 4.999863364711654e-05,
|
| 1661 |
+
"loss": 7.143423461914063,
|
| 1662 |
+
"step": 4700
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 0.05002647499272435,
|
| 1666 |
+
"grad_norm": 6.923771858215332,
|
| 1667 |
+
"learning_rate": 4.999856313779693e-05,
|
| 1668 |
+
"loss": 7.110890197753906,
|
| 1669 |
+
"step": 4720
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 0.1004768587909744,
|
| 1673 |
+
"grad_norm": 7.3437275886535645,
|
| 1674 |
+
"learning_rate": 4.999396360110861e-05,
|
| 1675 |
+
"loss": 9.684793260362413,
|
| 1676 |
+
"step": 4740
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 0.10090081178165361,
|
| 1680 |
+
"grad_norm": 8.959833145141602,
|
| 1681 |
+
"learning_rate": 4.999366739216703e-05,
|
| 1682 |
+
"loss": 9.575524139404298,
|
| 1683 |
+
"step": 4760
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 0.10132476477233283,
|
| 1687 |
+
"grad_norm": 11.047332763671875,
|
| 1688 |
+
"learning_rate": 4.999336408961287e-05,
|
| 1689 |
+
"loss": 9.669027709960938,
|
| 1690 |
+
"step": 4780
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 0.10174871776301204,
|
| 1694 |
+
"grad_norm": 8.867971420288086,
|
| 1695 |
+
"learning_rate": 4.999305369353221e-05,
|
| 1696 |
+
"loss": 9.227397155761718,
|
| 1697 |
+
"step": 4800
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 0.10217267075369126,
|
| 1701 |
+
"grad_norm": 10.266776084899902,
|
| 1702 |
+
"learning_rate": 4.999273620401315e-05,
|
| 1703 |
+
"loss": 10.01598892211914,
|
| 1704 |
+
"step": 4820
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 0.10259662374437048,
|
| 1708 |
+
"grad_norm": 8.681143760681152,
|
| 1709 |
+
"learning_rate": 4.9992411621145804e-05,
|
| 1710 |
+
"loss": 8.92789306640625,
|
| 1711 |
+
"step": 4840
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 0.10302057673504969,
|
| 1715 |
+
"grad_norm": 8.316998481750488,
|
| 1716 |
+
"learning_rate": 4.9992079945022273e-05,
|
| 1717 |
+
"loss": 9.240599822998046,
|
| 1718 |
+
"step": 4860
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 0.10344452972572891,
|
| 1722 |
+
"grad_norm": 7.769136905670166,
|
| 1723 |
+
"learning_rate": 4.9991741175736714e-05,
|
| 1724 |
+
"loss": 8.822511291503906,
|
| 1725 |
+
"step": 4880
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 0.10386848271640813,
|
| 1729 |
+
"grad_norm": 11.900764465332031,
|
| 1730 |
+
"learning_rate": 4.999139531338527e-05,
|
| 1731 |
+
"loss": 8.96667251586914,
|
| 1732 |
+
"step": 4900
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 0.10429243570708735,
|
| 1736 |
+
"grad_norm": 8.078374862670898,
|
| 1737 |
+
"learning_rate": 4.99910423580661e-05,
|
| 1738 |
+
"loss": 8.887445068359375,
|
| 1739 |
+
"step": 4920
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 0.10471638869776656,
|
| 1743 |
+
"grad_norm": 7.878390312194824,
|
| 1744 |
+
"learning_rate": 4.999068230987937e-05,
|
| 1745 |
+
"loss": 8.875336456298829,
|
| 1746 |
+
"step": 4940
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 0.10514034168844578,
|
| 1750 |
+
"grad_norm": 8.94696044921875,
|
| 1751 |
+
"learning_rate": 4.999031516892728e-05,
|
| 1752 |
+
"loss": 9.244612121582032,
|
| 1753 |
+
"step": 4960
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 0.105564294679125,
|
| 1757 |
+
"grad_norm": 7.411794662475586,
|
| 1758 |
+
"learning_rate": 4.9989940935314024e-05,
|
| 1759 |
+
"loss": 8.901412963867188,
|
| 1760 |
+
"step": 4980
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 0.10598824766980422,
|
| 1764 |
+
"grad_norm": 8.138178825378418,
|
| 1765 |
+
"learning_rate": 4.998955960914582e-05,
|
| 1766 |
+
"loss": 8.586210632324219,
|
| 1767 |
+
"step": 5000
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 0.10641220066048343,
|
| 1771 |
+
"grad_norm": 8.887142181396484,
|
| 1772 |
+
"learning_rate": 4.998917119053089e-05,
|
| 1773 |
+
"loss": 8.020191192626953,
|
| 1774 |
+
"step": 5020
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 0.10683615365116264,
|
| 1778 |
+
"grad_norm": 12.754515647888184,
|
| 1779 |
+
"learning_rate": 4.998877567957948e-05,
|
| 1780 |
+
"loss": 7.986856842041016,
|
| 1781 |
+
"step": 5040
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 0.10726010664184187,
|
| 1785 |
+
"grad_norm": 7.767778396606445,
|
| 1786 |
+
"learning_rate": 4.998837307640383e-05,
|
| 1787 |
+
"loss": 7.962983703613281,
|
| 1788 |
+
"step": 5060
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 0.10768405963252108,
|
| 1792 |
+
"grad_norm": 8.013529777526855,
|
| 1793 |
+
"learning_rate": 4.998796338111822e-05,
|
| 1794 |
+
"loss": 8.206927490234374,
|
| 1795 |
+
"step": 5080
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 0.1081080126232003,
|
| 1799 |
+
"grad_norm": 8.06114387512207,
|
| 1800 |
+
"learning_rate": 4.9987546593838924e-05,
|
| 1801 |
+
"loss": 8.40353775024414,
|
| 1802 |
+
"step": 5100
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 0.10853196561387951,
|
| 1806 |
+
"grad_norm": 8.638311386108398,
|
| 1807 |
+
"learning_rate": 4.9987122714684234e-05,
|
| 1808 |
+
"loss": 7.984647369384765,
|
| 1809 |
+
"step": 5120
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 0.10895591860455874,
|
| 1813 |
+
"grad_norm": 8.485541343688965,
|
| 1814 |
+
"learning_rate": 4.9986691743774436e-05,
|
| 1815 |
+
"loss": 7.266371154785157,
|
| 1816 |
+
"step": 5140
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 0.10937987159523795,
|
| 1820 |
+
"grad_norm": 6.6758928298950195,
|
| 1821 |
+
"learning_rate": 4.9986253681231876e-05,
|
| 1822 |
+
"loss": 7.941251373291015,
|
| 1823 |
+
"step": 5160
|
| 1824 |
+
},
|
| 1825 |
+
{
|
| 1826 |
+
"epoch": 0.10980382458591717,
|
| 1827 |
+
"grad_norm": 8.448532104492188,
|
| 1828 |
+
"learning_rate": 4.9985808527180865e-05,
|
| 1829 |
+
"loss": 7.619065856933593,
|
| 1830 |
+
"step": 5180
|
| 1831 |
+
},
|
| 1832 |
+
{
|
| 1833 |
+
"epoch": 0.11022777757659638,
|
| 1834 |
+
"grad_norm": 7.42982816696167,
|
| 1835 |
+
"learning_rate": 4.9985356281747746e-05,
|
| 1836 |
+
"loss": 7.369577789306641,
|
| 1837 |
+
"step": 5200
|
| 1838 |
+
},
|
| 1839 |
+
{
|
| 1840 |
+
"epoch": 0.1106517305672756,
|
| 1841 |
+
"grad_norm": 6.739811420440674,
|
| 1842 |
+
"learning_rate": 4.9984896945060875e-05,
|
| 1843 |
+
"loss": 7.164421081542969,
|
| 1844 |
+
"step": 5220
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 0.11107568355795482,
|
| 1848 |
+
"grad_norm": 8.617021560668945,
|
| 1849 |
+
"learning_rate": 4.998443051725063e-05,
|
| 1850 |
+
"loss": 7.638619995117187,
|
| 1851 |
+
"step": 5240
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 0.11149963654863403,
|
| 1855 |
+
"grad_norm": 7.801265716552734,
|
| 1856 |
+
"learning_rate": 4.998395699844938e-05,
|
| 1857 |
+
"loss": 7.813314056396484,
|
| 1858 |
+
"step": 5260
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 0.11192358953931325,
|
| 1862 |
+
"grad_norm": 7.472166061401367,
|
| 1863 |
+
"learning_rate": 4.998347638879152e-05,
|
| 1864 |
+
"loss": 7.440145874023438,
|
| 1865 |
+
"step": 5280
|
| 1866 |
+
},
|
| 1867 |
+
{
|
| 1868 |
+
"epoch": 0.11234754252999246,
|
| 1869 |
+
"grad_norm": 7.722754955291748,
|
| 1870 |
+
"learning_rate": 4.9982988688413444e-05,
|
| 1871 |
+
"loss": 7.479670715332031,
|
| 1872 |
+
"step": 5300
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 0.11277149552067169,
|
| 1876 |
+
"grad_norm": 6.163559436798096,
|
| 1877 |
+
"learning_rate": 4.9982493897453585e-05,
|
| 1878 |
+
"loss": 7.347272491455078,
|
| 1879 |
+
"step": 5320
|
| 1880 |
+
},
|
| 1881 |
+
{
|
| 1882 |
+
"epoch": 0.1131954485113509,
|
| 1883 |
+
"grad_norm": 10.241913795471191,
|
| 1884 |
+
"learning_rate": 4.998199201605237e-05,
|
| 1885 |
+
"loss": 7.39600830078125,
|
| 1886 |
+
"step": 5340
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 0.11361940150203012,
|
| 1890 |
+
"grad_norm": 7.981228351593018,
|
| 1891 |
+
"learning_rate": 4.9981483044352233e-05,
|
| 1892 |
+
"loss": 7.175743103027344,
|
| 1893 |
+
"step": 5360
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 0.11404335449270933,
|
| 1897 |
+
"grad_norm": 8.812307357788086,
|
| 1898 |
+
"learning_rate": 4.9980966982497634e-05,
|
| 1899 |
+
"loss": 7.342212677001953,
|
| 1900 |
+
"step": 5380
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"epoch": 0.11446730748338856,
|
| 1904 |
+
"grad_norm": 9.753350257873535,
|
| 1905 |
+
"learning_rate": 4.998044383063505e-05,
|
| 1906 |
+
"loss": 7.329566192626953,
|
| 1907 |
+
"step": 5400
|
| 1908 |
+
},
|
| 1909 |
+
{
|
| 1910 |
+
"epoch": 0.11489126047406777,
|
| 1911 |
+
"grad_norm": 7.825401782989502,
|
| 1912 |
+
"learning_rate": 4.997991358891294e-05,
|
| 1913 |
+
"loss": 7.079512023925782,
|
| 1914 |
+
"step": 5420
|
| 1915 |
+
},
|
| 1916 |
+
{
|
| 1917 |
+
"epoch": 0.11531521346474699,
|
| 1918 |
+
"grad_norm": 8.922019958496094,
|
| 1919 |
+
"learning_rate": 4.99793762574818e-05,
|
| 1920 |
+
"loss": 6.7873291015625,
|
| 1921 |
+
"step": 5440
|
| 1922 |
+
},
|
| 1923 |
+
{
|
| 1924 |
+
"epoch": 0.1157391664554262,
|
| 1925 |
+
"grad_norm": 10.937968254089355,
|
| 1926 |
+
"learning_rate": 4.997883183649416e-05,
|
| 1927 |
+
"loss": 6.742900085449219,
|
| 1928 |
+
"step": 5460
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 0.11616311944610541,
|
| 1932 |
+
"grad_norm": 9.819665908813477,
|
| 1933 |
+
"learning_rate": 4.997830806999486e-05,
|
| 1934 |
+
"loss": 7.3204811096191404,
|
| 1935 |
+
"step": 5480
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 0.11658707243678464,
|
| 1939 |
+
"grad_norm": 8.020406723022461,
|
| 1940 |
+
"learning_rate": 4.9977749824818256e-05,
|
| 1941 |
+
"loss": 6.910675811767578,
|
| 1942 |
+
"step": 5500
|
| 1943 |
+
},
|
| 1944 |
+
{
|
| 1945 |
+
"epoch": 0.11701102542746385,
|
| 1946 |
+
"grad_norm": 7.209946632385254,
|
| 1947 |
+
"learning_rate": 4.997718449054673e-05,
|
| 1948 |
+
"loss": 7.582695770263672,
|
| 1949 |
+
"step": 5520
|
| 1950 |
+
},
|
| 1951 |
+
{
|
| 1952 |
+
"epoch": 0.11743497841814307,
|
| 1953 |
+
"grad_norm": 6.4485883712768555,
|
| 1954 |
+
"learning_rate": 4.997661206734074e-05,
|
| 1955 |
+
"loss": 6.508930206298828,
|
| 1956 |
+
"step": 5540
|
| 1957 |
+
},
|
| 1958 |
+
{
|
| 1959 |
+
"epoch": 0.11785893140882228,
|
| 1960 |
+
"grad_norm": 17.4251651763916,
|
| 1961 |
+
"learning_rate": 4.9976032555362754e-05,
|
| 1962 |
+
"loss": 6.748028564453125,
|
| 1963 |
+
"step": 5560
|
| 1964 |
+
},
|
| 1965 |
+
{
|
| 1966 |
+
"epoch": 0.1182828843995015,
|
| 1967 |
+
"grad_norm": 9.216353416442871,
|
| 1968 |
+
"learning_rate": 4.9975445954777245e-05,
|
| 1969 |
+
"loss": 6.938046264648437,
|
| 1970 |
+
"step": 5580
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 0.11870683739018072,
|
| 1974 |
+
"grad_norm": 8.369783401489258,
|
| 1975 |
+
"learning_rate": 4.9974852265750695e-05,
|
| 1976 |
+
"loss": 6.490337371826172,
|
| 1977 |
+
"step": 5600
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 0.11913079038085994,
|
| 1981 |
+
"grad_norm": 7.427968978881836,
|
| 1982 |
+
"learning_rate": 4.9974251488451605e-05,
|
| 1983 |
+
"loss": 6.322509384155273,
|
| 1984 |
+
"step": 5620
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 0.11955474337153915,
|
| 1988 |
+
"grad_norm": 8.51576042175293,
|
| 1989 |
+
"learning_rate": 4.997364362305049e-05,
|
| 1990 |
+
"loss": 6.911622619628906,
|
| 1991 |
+
"step": 5640
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"epoch": 0.11997869636221838,
|
| 1995 |
+
"grad_norm": 6.536707401275635,
|
| 1996 |
+
"learning_rate": 4.9973028669719875e-05,
|
| 1997 |
+
"loss": 6.348305511474609,
|
| 1998 |
+
"step": 5660
|
| 1999 |
+
},
|
| 2000 |
+
{
|
| 2001 |
+
"epoch": 0.12040264935289759,
|
| 2002 |
+
"grad_norm": 5.895622253417969,
|
| 2003 |
+
"learning_rate": 4.997240662863428e-05,
|
| 2004 |
+
"loss": 6.29071273803711,
|
| 2005 |
+
"step": 5680
|
| 2006 |
+
},
|
| 2007 |
+
{
|
| 2008 |
+
"epoch": 0.1208266023435768,
|
| 2009 |
+
"grad_norm": 7.713855266571045,
|
| 2010 |
+
"learning_rate": 4.997177749997026e-05,
|
| 2011 |
+
"loss": 6.9540153503417965,
|
| 2012 |
+
"step": 5700
|
| 2013 |
+
},
|
| 2014 |
+
{
|
| 2015 |
+
"epoch": 0.12125055533425602,
|
| 2016 |
+
"grad_norm": 7.833205699920654,
|
| 2017 |
+
"learning_rate": 4.9971141283906374e-05,
|
| 2018 |
+
"loss": 6.268700790405274,
|
| 2019 |
+
"step": 5720
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 0.12167450832493523,
|
| 2023 |
+
"grad_norm": 8.855192184448242,
|
| 2024 |
+
"learning_rate": 4.9970497980623186e-05,
|
| 2025 |
+
"loss": 6.385813140869141,
|
| 2026 |
+
"step": 5740
|
| 2027 |
+
},
|
| 2028 |
+
{
|
| 2029 |
+
"epoch": 0.12209846131561446,
|
| 2030 |
+
"grad_norm": 7.562361240386963,
|
| 2031 |
+
"learning_rate": 4.996984759030328e-05,
|
| 2032 |
+
"loss": 6.241614532470703,
|
| 2033 |
+
"step": 5760
|
| 2034 |
+
},
|
| 2035 |
+
{
|
| 2036 |
+
"epoch": 0.12252241430629367,
|
| 2037 |
+
"grad_norm": 6.7725138664245605,
|
| 2038 |
+
"learning_rate": 4.9969190113131233e-05,
|
| 2039 |
+
"loss": 6.159543991088867,
|
| 2040 |
+
"step": 5780
|
| 2041 |
+
},
|
| 2042 |
+
{
|
| 2043 |
+
"epoch": 0.12294636729697289,
|
| 2044 |
+
"grad_norm": 6.906131267547607,
|
| 2045 |
+
"learning_rate": 4.996852554929367e-05,
|
| 2046 |
+
"loss": 6.404425048828125,
|
| 2047 |
+
"step": 5800
|
| 2048 |
+
},
|
| 2049 |
+
{
|
| 2050 |
+
"epoch": 0.1233703202876521,
|
| 2051 |
+
"grad_norm": 7.0016961097717285,
|
| 2052 |
+
"learning_rate": 4.996785389897919e-05,
|
| 2053 |
+
"loss": 6.4215339660644535,
|
| 2054 |
+
"step": 5820
|
| 2055 |
+
},
|
| 2056 |
+
{
|
| 2057 |
+
"epoch": 0.12379427327833133,
|
| 2058 |
+
"grad_norm": 8.028460502624512,
|
| 2059 |
+
"learning_rate": 4.996717516237843e-05,
|
| 2060 |
+
"loss": 6.3926044464111325,
|
| 2061 |
+
"step": 5840
|
| 2062 |
+
},
|
| 2063 |
+
{
|
| 2064 |
+
"epoch": 0.12421822626901054,
|
| 2065 |
+
"grad_norm": 13.41717529296875,
|
| 2066 |
+
"learning_rate": 4.996648933968401e-05,
|
| 2067 |
+
"loss": 6.214765548706055,
|
| 2068 |
+
"step": 5860
|
| 2069 |
+
},
|
| 2070 |
+
{
|
| 2071 |
+
"epoch": 0.12464217925968976,
|
| 2072 |
+
"grad_norm": 7.311347007751465,
|
| 2073 |
+
"learning_rate": 4.996579643109059e-05,
|
| 2074 |
+
"loss": 6.153236389160156,
|
| 2075 |
+
"step": 5880
|
| 2076 |
+
},
|
| 2077 |
+
{
|
| 2078 |
+
"epoch": 0.12506613225036897,
|
| 2079 |
+
"grad_norm": 6.1434526443481445,
|
| 2080 |
+
"learning_rate": 4.996509643679483e-05,
|
| 2081 |
+
"loss": 6.312967681884766,
|
| 2082 |
+
"step": 5900
|
| 2083 |
+
},
|
| 2084 |
+
{
|
| 2085 |
+
"epoch": 0.12549008524104818,
|
| 2086 |
+
"grad_norm": 10.802331924438477,
|
| 2087 |
+
"learning_rate": 4.996438935699539e-05,
|
| 2088 |
+
"loss": 6.46864013671875,
|
| 2089 |
+
"step": 5920
|
| 2090 |
+
},
|
| 2091 |
+
{
|
| 2092 |
+
"epoch": 0.1259140382317274,
|
| 2093 |
+
"grad_norm": 8.637312889099121,
|
| 2094 |
+
"learning_rate": 4.996367519189296e-05,
|
| 2095 |
+
"loss": 6.12041015625,
|
| 2096 |
+
"step": 5940
|
| 2097 |
+
},
|
| 2098 |
+
{
|
| 2099 |
+
"epoch": 0.12633799122240663,
|
| 2100 |
+
"grad_norm": 6.246335506439209,
|
| 2101 |
+
"learning_rate": 4.996295394169023e-05,
|
| 2102 |
+
"loss": 6.266645812988282,
|
| 2103 |
+
"step": 5960
|
| 2104 |
+
},
|
| 2105 |
+
{
|
| 2106 |
+
"epoch": 0.12676194421308584,
|
| 2107 |
+
"grad_norm": 6.469450950622559,
|
| 2108 |
+
"learning_rate": 4.9962225606591894e-05,
|
| 2109 |
+
"loss": 5.881629943847656,
|
| 2110 |
+
"step": 5980
|
| 2111 |
+
},
|
| 2112 |
+
{
|
| 2113 |
+
"epoch": 0.12718589720376505,
|
| 2114 |
+
"grad_norm": 7.862303733825684,
|
| 2115 |
+
"learning_rate": 4.9961490186804685e-05,
|
| 2116 |
+
"loss": 5.996078872680664,
|
| 2117 |
+
"step": 6000
|
| 2118 |
+
},
|
| 2119 |
+
{
|
| 2120 |
+
"epoch": 0.12760985019444426,
|
| 2121 |
+
"grad_norm": 6.547665596008301,
|
| 2122 |
+
"learning_rate": 4.99607476825373e-05,
|
| 2123 |
+
"loss": 6.404113006591797,
|
| 2124 |
+
"step": 6020
|
| 2125 |
+
}
|
| 2126 |
+
],
|
| 2127 |
+
"logging_steps": 20,
|
| 2128 |
+
"max_steps": 117940,
|
| 2129 |
+
"num_input_tokens_seen": 0,
|
| 2130 |
+
"num_train_epochs": 3,
|
| 2131 |
+
"save_steps": 1000000000,
|
| 2132 |
+
"stateful_callbacks": {
|
| 2133 |
+
"TrainerControl": {
|
| 2134 |
+
"args": {
|
| 2135 |
+
"should_epoch_stop": false,
|
| 2136 |
+
"should_evaluate": false,
|
| 2137 |
+
"should_log": false,
|
| 2138 |
+
"should_save": true,
|
| 2139 |
+
"should_training_stop": false
|
| 2140 |
+
},
|
| 2141 |
+
"attributes": {}
|
| 2142 |
+
}
|
| 2143 |
+
},
|
| 2144 |
+
"total_flos": 3.4565610901008384e+16,
|
| 2145 |
+
"train_batch_size": 1,
|
| 2146 |
+
"trial_name": null,
|
| 2147 |
+
"trial_params": null
|
| 2148 |
+
}
|