Upload folder using huggingface_hub
Browse files
slots/8/checkpoint-7967/config.json
ADDED
|
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"IsaiahForCausalLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_dropout": 0.0,
|
| 6 |
+
"dtype": "float32",
|
| 7 |
+
"hidden_dropout": 0.0,
|
| 8 |
+
"hidden_size": 1024,
|
| 9 |
+
"initializer_range": 0.02,
|
| 10 |
+
"intermediate_size": 4096,
|
| 11 |
+
"max_position_embeddings": 160,
|
| 12 |
+
"model_type": "isaiah",
|
| 13 |
+
"neftune_alpha": 0.0,
|
| 14 |
+
"num_attention_heads": 16,
|
| 15 |
+
"num_hidden_layers": 16,
|
| 16 |
+
"num_key_value_heads": 4,
|
| 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-7967/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d1d22b5fbb2f80cd05c16a315c8bc609f9c13aded2e79eaee760d8babbe26b5b
|
| 3 |
+
size 1302821336
|
slots/8/checkpoint-7967/trainer_state.json
ADDED
|
@@ -0,0 +1,2827 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.30241546619577026,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 7967,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.00015183386844227941,
|
| 14 |
+
"grad_norm": 91.41053771972656,
|
| 15 |
+
"learning_rate": 5e-05,
|
| 16 |
+
"loss": 1013.596923828125,
|
| 17 |
+
"step": 1
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.003036677368845588,
|
| 21 |
+
"grad_norm": 128.77392578125,
|
| 22 |
+
"learning_rate": 4.99998526084367e-05,
|
| 23 |
+
"loss": 928.9828330592105,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.006073354737691176,
|
| 28 |
+
"grad_norm": 89.98699951171875,
|
| 29 |
+
"learning_rate": 4.9999410435484726e-05,
|
| 30 |
+
"loss": 811.4373046875,
|
| 31 |
+
"step": 40
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.009110032106536764,
|
| 35 |
+
"grad_norm": 52.5091438293457,
|
| 36 |
+
"learning_rate": 4.9998573407405316e-05,
|
| 37 |
+
"loss": 761.7615234375,
|
| 38 |
+
"step": 60
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.012146709475382352,
|
| 42 |
+
"grad_norm": 43.218017578125,
|
| 43 |
+
"learning_rate": 4.999737247039103e-05,
|
| 44 |
+
"loss": 741.289501953125,
|
| 45 |
+
"step": 80
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.015183386844227941,
|
| 49 |
+
"grad_norm": 53.97433853149414,
|
| 50 |
+
"learning_rate": 4.9995807641924135e-05,
|
| 51 |
+
"loss": 725.4333984375,
|
| 52 |
+
"step": 100
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.01822006421307353,
|
| 56 |
+
"grad_norm": 58.722232818603516,
|
| 57 |
+
"learning_rate": 4.9993878944784136e-05,
|
| 58 |
+
"loss": 715.616015625,
|
| 59 |
+
"step": 120
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.021256741581919116,
|
| 63 |
+
"grad_norm": 63.764610290527344,
|
| 64 |
+
"learning_rate": 4.999158640704742e-05,
|
| 65 |
+
"loss": 698.364794921875,
|
| 66 |
+
"step": 140
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.024293418950764704,
|
| 70 |
+
"grad_norm": 81.58174896240234,
|
| 71 |
+
"learning_rate": 4.998893006208691e-05,
|
| 72 |
+
"loss": 685.3453125,
|
| 73 |
+
"step": 160
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.027330096319610295,
|
| 77 |
+
"grad_norm": 75.00901794433594,
|
| 78 |
+
"learning_rate": 4.9985909948571527e-05,
|
| 79 |
+
"loss": 675.01171875,
|
| 80 |
+
"step": 180
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.030366773688455882,
|
| 84 |
+
"grad_norm": 106.4965591430664,
|
| 85 |
+
"learning_rate": 4.998252611046561e-05,
|
| 86 |
+
"loss": 661.97890625,
|
| 87 |
+
"step": 200
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.03340345105730147,
|
| 91 |
+
"grad_norm": 66.66131591796875,
|
| 92 |
+
"learning_rate": 4.997877859702834e-05,
|
| 93 |
+
"loss": 656.54736328125,
|
| 94 |
+
"step": 220
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.03644012842614706,
|
| 98 |
+
"grad_norm": 100.64048767089844,
|
| 99 |
+
"learning_rate": 4.9974667462812965e-05,
|
| 100 |
+
"loss": 650.53447265625,
|
| 101 |
+
"step": 240
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.039476805794992645,
|
| 105 |
+
"grad_norm": 76.53395080566406,
|
| 106 |
+
"learning_rate": 4.9970192767666044e-05,
|
| 107 |
+
"loss": 639.15849609375,
|
| 108 |
+
"step": 260
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.04251348316383823,
|
| 112 |
+
"grad_norm": 88.33858489990234,
|
| 113 |
+
"learning_rate": 4.996535457672656e-05,
|
| 114 |
+
"loss": 633.5654296875,
|
| 115 |
+
"step": 280
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.04555016053268382,
|
| 119 |
+
"grad_norm": 84.0911865234375,
|
| 120 |
+
"learning_rate": 4.996015296042497e-05,
|
| 121 |
+
"loss": 634.099951171875,
|
| 122 |
+
"step": 300
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.04858691474672964,
|
| 126 |
+
"grad_norm": 94.41912078857422,
|
| 127 |
+
"learning_rate": 4.9999897644716974e-05,
|
| 128 |
+
"loss": 711.8211167279412,
|
| 129 |
+
"step": 320
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.05162359691840024,
|
| 133 |
+
"grad_norm": 91.57540893554688,
|
| 134 |
+
"learning_rate": 4.999944273403803e-05,
|
| 135 |
+
"loss": 691.16796875,
|
| 136 |
+
"step": 340
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.05466027909007084,
|
| 140 |
+
"grad_norm": 102.95606994628906,
|
| 141 |
+
"learning_rate": 4.9998673486357894e-05,
|
| 142 |
+
"loss": 684.7939453125,
|
| 143 |
+
"step": 360
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.05769696126174144,
|
| 147 |
+
"grad_norm": 99.65682983398438,
|
| 148 |
+
"learning_rate": 4.999750893953781e-05,
|
| 149 |
+
"loss": 680.48828125,
|
| 150 |
+
"step": 380
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.060733643433412046,
|
| 154 |
+
"grad_norm": 112.4695816040039,
|
| 155 |
+
"learning_rate": 4.9995980499278494e-05,
|
| 156 |
+
"loss": 675.817822265625,
|
| 157 |
+
"step": 400
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.06377032560508265,
|
| 161 |
+
"grad_norm": 82.0074691772461,
|
| 162 |
+
"learning_rate": 4.9994088187829756e-05,
|
| 163 |
+
"loss": 663.119140625,
|
| 164 |
+
"step": 420
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.06680700777675325,
|
| 168 |
+
"grad_norm": 96.62723541259766,
|
| 169 |
+
"learning_rate": 4.999183203273833e-05,
|
| 170 |
+
"loss": 661.77353515625,
|
| 171 |
+
"step": 440
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.06984368994842385,
|
| 175 |
+
"grad_norm": 99.35919952392578,
|
| 176 |
+
"learning_rate": 4.998921206684748e-05,
|
| 177 |
+
"loss": 656.063134765625,
|
| 178 |
+
"step": 460
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.07288037212009445,
|
| 182 |
+
"grad_norm": 97.6092758178711,
|
| 183 |
+
"learning_rate": 4.998622832829655e-05,
|
| 184 |
+
"loss": 658.578466796875,
|
| 185 |
+
"step": 480
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.07591705429176505,
|
| 189 |
+
"grad_norm": 88.02263641357422,
|
| 190 |
+
"learning_rate": 4.9982880860520386e-05,
|
| 191 |
+
"loss": 656.30810546875,
|
| 192 |
+
"step": 500
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.07895373646343566,
|
| 196 |
+
"grad_norm": 117.71761322021484,
|
| 197 |
+
"learning_rate": 4.99791697122487e-05,
|
| 198 |
+
"loss": 652.0546875,
|
| 199 |
+
"step": 520
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.023425828602828273,
|
| 203 |
+
"grad_norm": 114.52953338623047,
|
| 204 |
+
"learning_rate": 4.9999993722319546e-05,
|
| 205 |
+
"loss": 680.5703776041667,
|
| 206 |
+
"step": 540
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.024293451884414504,
|
| 210 |
+
"grad_norm": 112.35067749023438,
|
| 211 |
+
"learning_rate": 4.999995954797362e-05,
|
| 212 |
+
"loss": 666.2451171875,
|
| 213 |
+
"step": 560
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.025161075166000735,
|
| 217 |
+
"grad_norm": 84.08537292480469,
|
| 218 |
+
"learning_rate": 4.999989565684691e-05,
|
| 219 |
+
"loss": 664.393359375,
|
| 220 |
+
"step": 580
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.026028698447586967,
|
| 224 |
+
"grad_norm": 130.15928649902344,
|
| 225 |
+
"learning_rate": 4.9999802049015365e-05,
|
| 226 |
+
"loss": 662.348291015625,
|
| 227 |
+
"step": 600
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.026896321729173202,
|
| 231 |
+
"grad_norm": 103.99536895751953,
|
| 232 |
+
"learning_rate": 4.999967872459026e-05,
|
| 233 |
+
"loss": 659.2578125,
|
| 234 |
+
"step": 620
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.027763945010759433,
|
| 238 |
+
"grad_norm": 101.37040710449219,
|
| 239 |
+
"learning_rate": 4.9999525683718174e-05,
|
| 240 |
+
"loss": 658.54638671875,
|
| 241 |
+
"step": 640
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.028631568292345665,
|
| 245 |
+
"grad_norm": 98.98956298828125,
|
| 246 |
+
"learning_rate": 4.999934292658104e-05,
|
| 247 |
+
"loss": 652.716845703125,
|
| 248 |
+
"step": 660
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.0688314262698175,
|
| 252 |
+
"grad_norm": 114.9826431274414,
|
| 253 |
+
"learning_rate": 4.999993448199386e-05,
|
| 254 |
+
"loss": 676.5150390625,
|
| 255 |
+
"step": 680
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.07085587998363567,
|
| 259 |
+
"grad_norm": 97.52400207519531,
|
| 260 |
+
"learning_rate": 4.999970800043822e-05,
|
| 261 |
+
"loss": 672.913720703125,
|
| 262 |
+
"step": 700
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.07288033369745382,
|
| 266 |
+
"grad_norm": 119.27461242675781,
|
| 267 |
+
"learning_rate": 4.9999319747933995e-05,
|
| 268 |
+
"loss": 668.961376953125,
|
| 269 |
+
"step": 720
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.07490478741127199,
|
| 273 |
+
"grad_norm": 108.82205963134766,
|
| 274 |
+
"learning_rate": 4.999876972699352e-05,
|
| 275 |
+
"loss": 665.568994140625,
|
| 276 |
+
"step": 740
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.07692924112509016,
|
| 280 |
+
"grad_norm": 143.65585327148438,
|
| 281 |
+
"learning_rate": 4.999805794117596e-05,
|
| 282 |
+
"loss": 661.4259765625,
|
| 283 |
+
"step": 760
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.07895369483890831,
|
| 287 |
+
"grad_norm": 101.48106384277344,
|
| 288 |
+
"learning_rate": 4.9997231914115064e-05,
|
| 289 |
+
"loss": 660.966015625,
|
| 290 |
+
"step": 780
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.08097814855272648,
|
| 294 |
+
"grad_norm": 118.77201843261719,
|
| 295 |
+
"learning_rate": 4.9996204700988264e-05,
|
| 296 |
+
"loss": 654.88798828125,
|
| 297 |
+
"step": 800
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.08300260226654464,
|
| 301 |
+
"grad_norm": 97.97103881835938,
|
| 302 |
+
"learning_rate": 4.9995015739582426e-05,
|
| 303 |
+
"loss": 655.855517578125,
|
| 304 |
+
"step": 820
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.0850270559803628,
|
| 308 |
+
"grad_norm": 112.38823699951172,
|
| 309 |
+
"learning_rate": 4.9993665037591234e-05,
|
| 310 |
+
"loss": 649.156201171875,
|
| 311 |
+
"step": 840
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.08705150969418096,
|
| 315 |
+
"grad_norm": 98.14862060546875,
|
| 316 |
+
"learning_rate": 4.9992152603754935e-05,
|
| 317 |
+
"loss": 652.241064453125,
|
| 318 |
+
"step": 860
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.08907596340799913,
|
| 322 |
+
"grad_norm": 107.33179473876953,
|
| 323 |
+
"learning_rate": 4.9990478447860365e-05,
|
| 324 |
+
"loss": 649.800537109375,
|
| 325 |
+
"step": 880
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.09110041712181728,
|
| 329 |
+
"grad_norm": 89.38690948486328,
|
| 330 |
+
"learning_rate": 4.998864258074086e-05,
|
| 331 |
+
"loss": 644.5462890625,
|
| 332 |
+
"step": 900
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.09312487083563545,
|
| 336 |
+
"grad_norm": Infinity,
|
| 337 |
+
"learning_rate": 4.9986645014276134e-05,
|
| 338 |
+
"loss": 645.86826171875,
|
| 339 |
+
"step": 920
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.09514932454945361,
|
| 343 |
+
"grad_norm": 125.2510757446289,
|
| 344 |
+
"learning_rate": 4.9984597563877464e-05,
|
| 345 |
+
"loss": 643.441357421875,
|
| 346 |
+
"step": 940
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.09717377826327177,
|
| 350 |
+
"grad_norm": Infinity,
|
| 351 |
+
"learning_rate": 5e-05,
|
| 352 |
+
"loss": 599.23876953125,
|
| 353 |
+
"step": 960
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.09919823197708993,
|
| 357 |
+
"grad_norm": 124.58181762695312,
|
| 358 |
+
"learning_rate": 4.999992700000298e-05,
|
| 359 |
+
"loss": 565.266455078125,
|
| 360 |
+
"step": 980
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.10122268569090809,
|
| 364 |
+
"grad_norm": 97.4952163696289,
|
| 365 |
+
"learning_rate": 4.999969242985639e-05,
|
| 366 |
+
"loss": 552.871435546875,
|
| 367 |
+
"step": 1000
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.10324713940472625,
|
| 371 |
+
"grad_norm": 91.78075408935547,
|
| 372 |
+
"learning_rate": 4.999929608886198e-05,
|
| 373 |
+
"loss": 553.4515625,
|
| 374 |
+
"step": 1020
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.10527159311854442,
|
| 378 |
+
"grad_norm": 90.69072723388672,
|
| 379 |
+
"learning_rate": 4.999876972699352e-05,
|
| 380 |
+
"loss": 545.360595703125,
|
| 381 |
+
"step": 1040
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.10729604683236257,
|
| 385 |
+
"grad_norm": 82.90885925292969,
|
| 386 |
+
"learning_rate": 4.999805794117596e-05,
|
| 387 |
+
"loss": 545.597119140625,
|
| 388 |
+
"step": 1060
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.10932050054618074,
|
| 392 |
+
"grad_norm": 109.15331268310547,
|
| 393 |
+
"learning_rate": 4.99971843950872e-05,
|
| 394 |
+
"loss": 540.0650390625,
|
| 395 |
+
"step": 1080
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.11134495425999891,
|
| 399 |
+
"grad_norm": 96.9873275756836,
|
| 400 |
+
"learning_rate": 4.999614909437988e-05,
|
| 401 |
+
"loss": 534.475830078125,
|
| 402 |
+
"step": 1100
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.11336940797381706,
|
| 406 |
+
"grad_norm": 86.17156982421875,
|
| 407 |
+
"learning_rate": 4.9994952045753366e-05,
|
| 408 |
+
"loss": 539.829296875,
|
| 409 |
+
"step": 1120
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.11539386168763523,
|
| 413 |
+
"grad_norm": 122.1972427368164,
|
| 414 |
+
"learning_rate": 4.999359325695363e-05,
|
| 415 |
+
"loss": 536.82841796875,
|
| 416 |
+
"step": 1140
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.1174183154014534,
|
| 420 |
+
"grad_norm": 80.08365631103516,
|
| 421 |
+
"learning_rate": 4.9992072736773275e-05,
|
| 422 |
+
"loss": 524.685400390625,
|
| 423 |
+
"step": 1160
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.11944276911527155,
|
| 427 |
+
"grad_norm": 85.38316345214844,
|
| 428 |
+
"learning_rate": 4.999039049505148e-05,
|
| 429 |
+
"loss": 527.760791015625,
|
| 430 |
+
"step": 1180
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.12146722282908971,
|
| 434 |
+
"grad_norm": 116.98209381103516,
|
| 435 |
+
"learning_rate": 4.998854654267387e-05,
|
| 436 |
+
"loss": 530.089306640625,
|
| 437 |
+
"step": 1200
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.12349167654290788,
|
| 441 |
+
"grad_norm": 81.59492492675781,
|
| 442 |
+
"learning_rate": 4.998654089157249e-05,
|
| 443 |
+
"loss": 524.530908203125,
|
| 444 |
+
"step": 1220
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.12551613025672603,
|
| 448 |
+
"grad_norm": 87.97933959960938,
|
| 449 |
+
"learning_rate": 4.998437355472576e-05,
|
| 450 |
+
"loss": 520.95693359375,
|
| 451 |
+
"step": 1240
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.12754058397054419,
|
| 455 |
+
"grad_norm": 88.46687316894531,
|
| 456 |
+
"learning_rate": 4.998204454615831e-05,
|
| 457 |
+
"loss": 517.149560546875,
|
| 458 |
+
"step": 1260
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.12956503768436237,
|
| 462 |
+
"grad_norm": 97.07532501220703,
|
| 463 |
+
"learning_rate": 4.997955388094097e-05,
|
| 464 |
+
"loss": 516.886376953125,
|
| 465 |
+
"step": 1280
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.13158949139818052,
|
| 469 |
+
"grad_norm": 105.70747375488281,
|
| 470 |
+
"learning_rate": 4.997690157519059e-05,
|
| 471 |
+
"loss": 517.590673828125,
|
| 472 |
+
"step": 1300
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.13361394511199867,
|
| 476 |
+
"grad_norm": 114.41730499267578,
|
| 477 |
+
"learning_rate": 4.997408764607005e-05,
|
| 478 |
+
"loss": 515.4822265625,
|
| 479 |
+
"step": 1320
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.13563839882581685,
|
| 483 |
+
"grad_norm": 86.80391693115234,
|
| 484 |
+
"learning_rate": 4.997111211178801e-05,
|
| 485 |
+
"loss": 512.03515625,
|
| 486 |
+
"step": 1340
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.137662852539635,
|
| 490 |
+
"grad_norm": 80.7691421508789,
|
| 491 |
+
"learning_rate": 4.996797499159892e-05,
|
| 492 |
+
"loss": 506.872998046875,
|
| 493 |
+
"step": 1360
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.13968730625345316,
|
| 497 |
+
"grad_norm": 91.66568756103516,
|
| 498 |
+
"learning_rate": 4.9964676305802794e-05,
|
| 499 |
+
"loss": 514.529248046875,
|
| 500 |
+
"step": 1380
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.14171175996727134,
|
| 504 |
+
"grad_norm": 89.29586791992188,
|
| 505 |
+
"learning_rate": 4.9961216075745145e-05,
|
| 506 |
+
"loss": 508.17666015625,
|
| 507 |
+
"step": 1400
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.1437362136810895,
|
| 511 |
+
"grad_norm": 109.5767593383789,
|
| 512 |
+
"learning_rate": 4.9957594323816834e-05,
|
| 513 |
+
"loss": 503.719482421875,
|
| 514 |
+
"step": 1420
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.14576066739490764,
|
| 518 |
+
"grad_norm": 88.2542495727539,
|
| 519 |
+
"learning_rate": 4.995381107345389e-05,
|
| 520 |
+
"loss": 499.80751953125,
|
| 521 |
+
"step": 1440
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.14778512110872583,
|
| 525 |
+
"grad_norm": 98.4202651977539,
|
| 526 |
+
"learning_rate": 4.9949866349137414e-05,
|
| 527 |
+
"loss": 497.896435546875,
|
| 528 |
+
"step": 1460
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.14980957482254398,
|
| 532 |
+
"grad_norm": 101.47189331054688,
|
| 533 |
+
"learning_rate": 4.994576017639336e-05,
|
| 534 |
+
"loss": 497.0865234375,
|
| 535 |
+
"step": 1480
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.15183402853636213,
|
| 539 |
+
"grad_norm": 99.08831787109375,
|
| 540 |
+
"learning_rate": 4.9941492581792434e-05,
|
| 541 |
+
"loss": 498.08486328125,
|
| 542 |
+
"step": 1500
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.1538584822501803,
|
| 546 |
+
"grad_norm": 90.61394500732422,
|
| 547 |
+
"learning_rate": 4.993706359294986e-05,
|
| 548 |
+
"loss": 492.168408203125,
|
| 549 |
+
"step": 1520
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.15588293596399846,
|
| 553 |
+
"grad_norm": 110.01075744628906,
|
| 554 |
+
"learning_rate": 4.9932473238525247e-05,
|
| 555 |
+
"loss": 495.432958984375,
|
| 556 |
+
"step": 1540
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.15790738967781662,
|
| 560 |
+
"grad_norm": 100.31326293945312,
|
| 561 |
+
"learning_rate": 4.9927721548222374e-05,
|
| 562 |
+
"loss": 488.998291015625,
|
| 563 |
+
"step": 1560
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.1599318433916348,
|
| 567 |
+
"grad_norm": 79.81046295166016,
|
| 568 |
+
"learning_rate": 4.992280855278903e-05,
|
| 569 |
+
"loss": 486.6220703125,
|
| 570 |
+
"step": 1580
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.16195629710545295,
|
| 574 |
+
"grad_norm": 103.98905944824219,
|
| 575 |
+
"learning_rate": 4.991773428401676e-05,
|
| 576 |
+
"loss": 486.88388671875,
|
| 577 |
+
"step": 1600
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.1639807508192711,
|
| 581 |
+
"grad_norm": 91.88626861572266,
|
| 582 |
+
"learning_rate": 4.991249877474075e-05,
|
| 583 |
+
"loss": 487.03994140625,
|
| 584 |
+
"step": 1620
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.16600520453308928,
|
| 588 |
+
"grad_norm": 99.99327850341797,
|
| 589 |
+
"learning_rate": 4.9907102058839504e-05,
|
| 590 |
+
"loss": 487.4203125,
|
| 591 |
+
"step": 1640
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.16802965824690744,
|
| 595 |
+
"grad_norm": 101.7360610961914,
|
| 596 |
+
"learning_rate": 4.990154417123472e-05,
|
| 597 |
+
"loss": 485.78876953125,
|
| 598 |
+
"step": 1660
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.1700541119607256,
|
| 602 |
+
"grad_norm": 84.13375091552734,
|
| 603 |
+
"learning_rate": 4.9895825147891e-05,
|
| 604 |
+
"loss": 486.884326171875,
|
| 605 |
+
"step": 1680
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.17207856567454377,
|
| 609 |
+
"grad_norm": 109.67373657226562,
|
| 610 |
+
"learning_rate": 4.988994502581565e-05,
|
| 611 |
+
"loss": 485.81005859375,
|
| 612 |
+
"step": 1700
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.17410301938836192,
|
| 616 |
+
"grad_norm": 83.98812103271484,
|
| 617 |
+
"learning_rate": 4.988390384305843e-05,
|
| 618 |
+
"loss": 483.68251953125,
|
| 619 |
+
"step": 1720
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.10567650614643256,
|
| 623 |
+
"grad_norm": Infinity,
|
| 624 |
+
"learning_rate": 5e-05,
|
| 625 |
+
"loss": 638.3512573242188,
|
| 626 |
+
"step": 1740
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.10689117863087431,
|
| 630 |
+
"grad_norm": 119.68628692626953,
|
| 631 |
+
"learning_rate": 4.9999976411601244e-05,
|
| 632 |
+
"loss": 612.46064453125,
|
| 633 |
+
"step": 1760
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.10810585111531607,
|
| 637 |
+
"grad_norm": 177.01918029785156,
|
| 638 |
+
"learning_rate": 4.999990033178544e-05,
|
| 639 |
+
"loss": 602.578076171875,
|
| 640 |
+
"step": 1780
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.10932052359975783,
|
| 644 |
+
"grad_norm": 171.78785705566406,
|
| 645 |
+
"learning_rate": 4.9999763461114946e-05,
|
| 646 |
+
"loss": 595.524169921875,
|
| 647 |
+
"step": 1800
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.11053519608419958,
|
| 651 |
+
"grad_norm": 127.83625793457031,
|
| 652 |
+
"learning_rate": 4.999956834803778e-05,
|
| 653 |
+
"loss": 593.985400390625,
|
| 654 |
+
"step": 1820
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.11174986856864133,
|
| 658 |
+
"grad_norm": 118.50973510742188,
|
| 659 |
+
"learning_rate": 4.9999314993008504e-05,
|
| 660 |
+
"loss": 593.784814453125,
|
| 661 |
+
"step": 1840
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.11296454105308308,
|
| 665 |
+
"grad_norm": 173.1148681640625,
|
| 666 |
+
"learning_rate": 4.999900339661735e-05,
|
| 667 |
+
"loss": 592.070361328125,
|
| 668 |
+
"step": 1860
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.11417921353752485,
|
| 672 |
+
"grad_norm": 136.09713745117188,
|
| 673 |
+
"learning_rate": 4.9998633559590266e-05,
|
| 674 |
+
"loss": 589.0634765625,
|
| 675 |
+
"step": 1880
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.1153938860219666,
|
| 679 |
+
"grad_norm": 122.50788879394531,
|
| 680 |
+
"learning_rate": 4.999820548278885e-05,
|
| 681 |
+
"loss": 582.531005859375,
|
| 682 |
+
"step": 1900
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.11660855850640835,
|
| 686 |
+
"grad_norm": 105.60628509521484,
|
| 687 |
+
"learning_rate": 4.999771916721042e-05,
|
| 688 |
+
"loss": 590.70634765625,
|
| 689 |
+
"step": 1920
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.1178232309908501,
|
| 693 |
+
"grad_norm": 111.0425796508789,
|
| 694 |
+
"learning_rate": 4.999717461398794e-05,
|
| 695 |
+
"loss": 581.921240234375,
|
| 696 |
+
"step": 1940
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.11903790347529185,
|
| 700 |
+
"grad_norm": 136.53347778320312,
|
| 701 |
+
"learning_rate": 4.9996571824390065e-05,
|
| 702 |
+
"loss": 577.843115234375,
|
| 703 |
+
"step": 1960
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.12025257595973361,
|
| 707 |
+
"grad_norm": 103.87166595458984,
|
| 708 |
+
"learning_rate": 4.9995910799821134e-05,
|
| 709 |
+
"loss": 581.935009765625,
|
| 710 |
+
"step": 1980
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.12146724844417536,
|
| 714 |
+
"grad_norm": 122.83851623535156,
|
| 715 |
+
"learning_rate": 4.999519154182114e-05,
|
| 716 |
+
"loss": 581.473779296875,
|
| 717 |
+
"step": 2000
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.30670489933873085,
|
| 721 |
+
"grad_norm": 118.4366455078125,
|
| 722 |
+
"learning_rate": 5e-05,
|
| 723 |
+
"loss": 566.3182373046875,
|
| 724 |
+
"step": 2020
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 0.30974158151040143,
|
| 728 |
+
"grad_norm": 93.9906997680664,
|
| 729 |
+
"learning_rate": 4.999983577670253e-05,
|
| 730 |
+
"loss": 572.564208984375,
|
| 731 |
+
"step": 2040
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.312778263682072,
|
| 735 |
+
"grad_norm": 117.0743637084961,
|
| 736 |
+
"learning_rate": 4.999934310896764e-05,
|
| 737 |
+
"loss": 563.7841796875,
|
| 738 |
+
"step": 2060
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.31581494585374265,
|
| 742 |
+
"grad_norm": 115.43646240234375,
|
| 743 |
+
"learning_rate": 4.999846968935902e-05,
|
| 744 |
+
"loss": 560.27802734375,
|
| 745 |
+
"step": 2080
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.3188516280254132,
|
| 749 |
+
"grad_norm": 114.64088439941406,
|
| 750 |
+
"learning_rate": 4.999723236232536e-05,
|
| 751 |
+
"loss": 561.4779296875,
|
| 752 |
+
"step": 2100
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.32188831019708386,
|
| 756 |
+
"grad_norm": 115.49713134765625,
|
| 757 |
+
"learning_rate": 4.9995631145878675e-05,
|
| 758 |
+
"loss": 569.9380859375,
|
| 759 |
+
"step": 2120
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.32492499236875444,
|
| 763 |
+
"grad_norm": 90.72515869140625,
|
| 764 |
+
"learning_rate": 4.999366606332817e-05,
|
| 765 |
+
"loss": 560.292333984375,
|
| 766 |
+
"step": 2140
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 0.327961674540425,
|
| 770 |
+
"grad_norm": 94.22215270996094,
|
| 771 |
+
"learning_rate": 4.999133714327992e-05,
|
| 772 |
+
"loss": 553.107666015625,
|
| 773 |
+
"step": 2160
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.33099835671209565,
|
| 777 |
+
"grad_norm": 102.5918960571289,
|
| 778 |
+
"learning_rate": 4.998864441963643e-05,
|
| 779 |
+
"loss": 556.195361328125,
|
| 780 |
+
"step": 2180
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 0.33403503888376623,
|
| 784 |
+
"grad_norm": 130.70706176757812,
|
| 785 |
+
"learning_rate": 4.998558793159622e-05,
|
| 786 |
+
"loss": 556.1953125,
|
| 787 |
+
"step": 2200
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 0.33707172105543687,
|
| 791 |
+
"grad_norm": 111.88673400878906,
|
| 792 |
+
"learning_rate": 4.998216772365315e-05,
|
| 793 |
+
"loss": 557.0876953125,
|
| 794 |
+
"step": 2220
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.34010840322710745,
|
| 798 |
+
"grad_norm": 117.71963500976562,
|
| 799 |
+
"learning_rate": 4.9978383845595826e-05,
|
| 800 |
+
"loss": 557.620068359375,
|
| 801 |
+
"step": 2240
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.343145085398778,
|
| 805 |
+
"grad_norm": 97.68840026855469,
|
| 806 |
+
"learning_rate": 4.9974236352506874e-05,
|
| 807 |
+
"loss": 555.032373046875,
|
| 808 |
+
"step": 2260
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 0.34618176757044866,
|
| 812 |
+
"grad_norm": 110.69097900390625,
|
| 813 |
+
"learning_rate": 4.996972530476213e-05,
|
| 814 |
+
"loss": 554.2564453125,
|
| 815 |
+
"step": 2280
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 0.34921844974211924,
|
| 819 |
+
"grad_norm": 117.8437728881836,
|
| 820 |
+
"learning_rate": 4.996485076802976e-05,
|
| 821 |
+
"loss": 553.471826171875,
|
| 822 |
+
"step": 2300
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 0.3522551319137899,
|
| 826 |
+
"grad_norm": 93.11846923828125,
|
| 827 |
+
"learning_rate": 4.995961281326932e-05,
|
| 828 |
+
"loss": 547.875341796875,
|
| 829 |
+
"step": 2320
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 0.35529181408546046,
|
| 833 |
+
"grad_norm": 112.87896728515625,
|
| 834 |
+
"learning_rate": 4.99540115167307e-05,
|
| 835 |
+
"loss": 555.697265625,
|
| 836 |
+
"step": 2340
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.3583284962571311,
|
| 840 |
+
"grad_norm": 97.52216339111328,
|
| 841 |
+
"learning_rate": 4.994804695995303e-05,
|
| 842 |
+
"loss": 546.91748046875,
|
| 843 |
+
"step": 2360
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 0.36136517842880167,
|
| 847 |
+
"grad_norm": 104.06327056884766,
|
| 848 |
+
"learning_rate": 4.994171922976348e-05,
|
| 849 |
+
"loss": 545.298828125,
|
| 850 |
+
"step": 2380
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 0.36440186060047225,
|
| 854 |
+
"grad_norm": 110.4811019897461,
|
| 855 |
+
"learning_rate": 4.993502841827601e-05,
|
| 856 |
+
"loss": 545.938232421875,
|
| 857 |
+
"step": 2400
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 0.3674385427721429,
|
| 861 |
+
"grad_norm": 94.96924591064453,
|
| 862 |
+
"learning_rate": 4.992797462289002e-05,
|
| 863 |
+
"loss": 543.59873046875,
|
| 864 |
+
"step": 2420
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 0.37047522494381346,
|
| 868 |
+
"grad_norm": 105.82535552978516,
|
| 869 |
+
"learning_rate": 4.9920557946288946e-05,
|
| 870 |
+
"loss": 546.985009765625,
|
| 871 |
+
"step": 2440
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 0.3735119071154841,
|
| 875 |
+
"grad_norm": 116.1602554321289,
|
| 876 |
+
"learning_rate": 4.991277849643873e-05,
|
| 877 |
+
"loss": 543.735498046875,
|
| 878 |
+
"step": 2460
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 0.3765485892871547,
|
| 882 |
+
"grad_norm": 114.22776794433594,
|
| 883 |
+
"learning_rate": 4.990463638658629e-05,
|
| 884 |
+
"loss": 541.33115234375,
|
| 885 |
+
"step": 2480
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 0.37958527145882526,
|
| 889 |
+
"grad_norm": 110.11952209472656,
|
| 890 |
+
"learning_rate": 4.989613173525786e-05,
|
| 891 |
+
"loss": 539.22998046875,
|
| 892 |
+
"step": 2500
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 0.3826219536304959,
|
| 896 |
+
"grad_norm": 100.76897430419922,
|
| 897 |
+
"learning_rate": 4.9887264666257225e-05,
|
| 898 |
+
"loss": 534.1103515625,
|
| 899 |
+
"step": 2520
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 0.38565863580216647,
|
| 903 |
+
"grad_norm": 143.88633728027344,
|
| 904 |
+
"learning_rate": 4.9878035308664005e-05,
|
| 905 |
+
"loss": 542.872705078125,
|
| 906 |
+
"step": 2540
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 0.3886953179738371,
|
| 910 |
+
"grad_norm": 141.91552734375,
|
| 911 |
+
"learning_rate": 4.9868443796831686e-05,
|
| 912 |
+
"loss": 540.77744140625,
|
| 913 |
+
"step": 2560
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 0.3917320001455077,
|
| 917 |
+
"grad_norm": 107.48712158203125,
|
| 918 |
+
"learning_rate": 4.985849027038571e-05,
|
| 919 |
+
"loss": 535.156103515625,
|
| 920 |
+
"step": 2580
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 0.39476868231717827,
|
| 924 |
+
"grad_norm": 108.82980346679688,
|
| 925 |
+
"learning_rate": 4.984817487422142e-05,
|
| 926 |
+
"loss": 537.34169921875,
|
| 927 |
+
"step": 2600
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 0.3978053644888489,
|
| 931 |
+
"grad_norm": 144.6007080078125,
|
| 932 |
+
"learning_rate": 4.983749775850199e-05,
|
| 933 |
+
"loss": 538.90009765625,
|
| 934 |
+
"step": 2620
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 0.4008420466605195,
|
| 938 |
+
"grad_norm": 137.44659423828125,
|
| 939 |
+
"learning_rate": 4.9826459078656194e-05,
|
| 940 |
+
"loss": 542.628076171875,
|
| 941 |
+
"step": 2640
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 0.4038787288321901,
|
| 945 |
+
"grad_norm": 115.24795532226562,
|
| 946 |
+
"learning_rate": 4.981505899537619e-05,
|
| 947 |
+
"loss": 533.2291015625,
|
| 948 |
+
"step": 2660
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 0.4069154110038607,
|
| 952 |
+
"grad_norm": 114.89422607421875,
|
| 953 |
+
"learning_rate": 4.980329767461511e-05,
|
| 954 |
+
"loss": 532.064208984375,
|
| 955 |
+
"step": 2680
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 0.4099520931755313,
|
| 959 |
+
"grad_norm": 134.90200805664062,
|
| 960 |
+
"learning_rate": 4.9791175287584725e-05,
|
| 961 |
+
"loss": 537.6201171875,
|
| 962 |
+
"step": 2700
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 0.4129887753472019,
|
| 966 |
+
"grad_norm": 133.92173767089844,
|
| 967 |
+
"learning_rate": 4.9778692010752903e-05,
|
| 968 |
+
"loss": 531.06826171875,
|
| 969 |
+
"step": 2720
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 0.4160254575188725,
|
| 973 |
+
"grad_norm": 104.6704330444336,
|
| 974 |
+
"learning_rate": 4.976584802584105e-05,
|
| 975 |
+
"loss": 536.9376953125,
|
| 976 |
+
"step": 2740
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 0.4190621396905431,
|
| 980 |
+
"grad_norm": 101.63923645019531,
|
| 981 |
+
"learning_rate": 4.9752643519821477e-05,
|
| 982 |
+
"loss": 536.770458984375,
|
| 983 |
+
"step": 2760
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 0.4220988218622137,
|
| 987 |
+
"grad_norm": 100.10745239257812,
|
| 988 |
+
"learning_rate": 4.9739078684914656e-05,
|
| 989 |
+
"loss": 536.27080078125,
|
| 990 |
+
"step": 2780
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 0.42513550403388434,
|
| 994 |
+
"grad_norm": 106.5391616821289,
|
| 995 |
+
"learning_rate": 4.9725153718586426e-05,
|
| 996 |
+
"loss": 529.2302734375,
|
| 997 |
+
"step": 2800
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 0.4281721862055549,
|
| 1001 |
+
"grad_norm": 117.57733154296875,
|
| 1002 |
+
"learning_rate": 4.971086882354516e-05,
|
| 1003 |
+
"loss": 527.634326171875,
|
| 1004 |
+
"step": 2820
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 0.4312088683772255,
|
| 1008 |
+
"grad_norm": 111.12049102783203,
|
| 1009 |
+
"learning_rate": 4.969622420773876e-05,
|
| 1010 |
+
"loss": 522.284521484375,
|
| 1011 |
+
"step": 2840
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 0.43424555054889613,
|
| 1015 |
+
"grad_norm": 120.52484893798828,
|
| 1016 |
+
"learning_rate": 4.968122008435166e-05,
|
| 1017 |
+
"loss": 530.725,
|
| 1018 |
+
"step": 2860
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 0.4372822327205667,
|
| 1022 |
+
"grad_norm": 99.64852142333984,
|
| 1023 |
+
"learning_rate": 4.966585667180168e-05,
|
| 1024 |
+
"loss": 524.964111328125,
|
| 1025 |
+
"step": 2880
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 0.44031891489223735,
|
| 1029 |
+
"grad_norm": 106.82322692871094,
|
| 1030 |
+
"learning_rate": 4.965013419373693e-05,
|
| 1031 |
+
"loss": 522.35517578125,
|
| 1032 |
+
"step": 2900
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 0.4433555970639079,
|
| 1036 |
+
"grad_norm": 87.97702026367188,
|
| 1037 |
+
"learning_rate": 4.963405287903246e-05,
|
| 1038 |
+
"loss": 528.875732421875,
|
| 1039 |
+
"step": 2920
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 0.4463922792355785,
|
| 1043 |
+
"grad_norm": 100.42816162109375,
|
| 1044 |
+
"learning_rate": 4.9617612961787016e-05,
|
| 1045 |
+
"loss": 528.40126953125,
|
| 1046 |
+
"step": 2940
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 0.44942896140724914,
|
| 1050 |
+
"grad_norm": 110.56478118896484,
|
| 1051 |
+
"learning_rate": 4.960081468131954e-05,
|
| 1052 |
+
"loss": 526.7787109375,
|
| 1053 |
+
"step": 2960
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 0.4524656435789197,
|
| 1057 |
+
"grad_norm": 112.1586685180664,
|
| 1058 |
+
"learning_rate": 4.958365828216579e-05,
|
| 1059 |
+
"loss": 526.75107421875,
|
| 1060 |
+
"step": 2980
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 0.45550232575059035,
|
| 1064 |
+
"grad_norm": 96.77159881591797,
|
| 1065 |
+
"learning_rate": 4.956614401407468e-05,
|
| 1066 |
+
"loss": 518.793896484375,
|
| 1067 |
+
"step": 3000
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 0.45853900792226093,
|
| 1071 |
+
"grad_norm": 104.14801025390625,
|
| 1072 |
+
"learning_rate": 4.954827213200469e-05,
|
| 1073 |
+
"loss": 523.8990234375,
|
| 1074 |
+
"step": 3020
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 0.4615756900939315,
|
| 1078 |
+
"grad_norm": 121.8068618774414,
|
| 1079 |
+
"learning_rate": 4.953004289612019e-05,
|
| 1080 |
+
"loss": 525.7564453125,
|
| 1081 |
+
"step": 3040
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 0.13274636208269353,
|
| 1085 |
+
"grad_norm": 126.11927795410156,
|
| 1086 |
+
"learning_rate": 4.9999999665685646e-05,
|
| 1087 |
+
"loss": 555.1658528645834,
|
| 1088 |
+
"step": 3060
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 0.13361398536427976,
|
| 1092 |
+
"grad_norm": 105.76728820800781,
|
| 1093 |
+
"learning_rate": 4.999998202131891e-05,
|
| 1094 |
+
"loss": 552.33125,
|
| 1095 |
+
"step": 3080
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 0.134481608645866,
|
| 1099 |
+
"grad_norm": 140.261962890625,
|
| 1100 |
+
"learning_rate": 4.999993447441449e-05,
|
| 1101 |
+
"loss": 545.12216796875,
|
| 1102 |
+
"step": 3100
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 0.13534923192745224,
|
| 1106 |
+
"grad_norm": 136.40731811523438,
|
| 1107 |
+
"learning_rate": 4.9999857210759095e-05,
|
| 1108 |
+
"loss": 544.97294921875,
|
| 1109 |
+
"step": 3120
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 0.13621685520903848,
|
| 1113 |
+
"grad_norm": 115.48565673828125,
|
| 1114 |
+
"learning_rate": 4.999975023044456e-05,
|
| 1115 |
+
"loss": 543.114306640625,
|
| 1116 |
+
"step": 3140
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 0.1370844784906247,
|
| 1120 |
+
"grad_norm": 144.88966369628906,
|
| 1121 |
+
"learning_rate": 4.9999613533598065e-05,
|
| 1122 |
+
"loss": 543.839453125,
|
| 1123 |
+
"step": 3160
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 0.13795210177221093,
|
| 1127 |
+
"grad_norm": 105.20838928222656,
|
| 1128 |
+
"learning_rate": 4.9999447120382094e-05,
|
| 1129 |
+
"loss": 545.475341796875,
|
| 1130 |
+
"step": 3180
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 0.13881972505379717,
|
| 1134 |
+
"grad_norm": 120.91262817382812,
|
| 1135 |
+
"learning_rate": 4.999925099099444e-05,
|
| 1136 |
+
"loss": 540.4140625,
|
| 1137 |
+
"step": 3200
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 0.1396873483353834,
|
| 1141 |
+
"grad_norm": 128.4713897705078,
|
| 1142 |
+
"learning_rate": 4.9999025145668256e-05,
|
| 1143 |
+
"loss": 543.10029296875,
|
| 1144 |
+
"step": 3220
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 0.14055497161696962,
|
| 1148 |
+
"grad_norm": 129.53460693359375,
|
| 1149 |
+
"learning_rate": 4.999876958467199e-05,
|
| 1150 |
+
"loss": 543.0662109375,
|
| 1151 |
+
"step": 3240
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 0.14142259489855585,
|
| 1155 |
+
"grad_norm": 110.85163116455078,
|
| 1156 |
+
"learning_rate": 4.9998484308309424e-05,
|
| 1157 |
+
"loss": 535.75615234375,
|
| 1158 |
+
"step": 3260
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 0.1422902181801421,
|
| 1162 |
+
"grad_norm": 135.63595581054688,
|
| 1163 |
+
"learning_rate": 4.9998169316919654e-05,
|
| 1164 |
+
"loss": 537.56318359375,
|
| 1165 |
+
"step": 3280
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 0.14315784146172833,
|
| 1169 |
+
"grad_norm": 141.77488708496094,
|
| 1170 |
+
"learning_rate": 4.999782461087711e-05,
|
| 1171 |
+
"loss": 530.75205078125,
|
| 1172 |
+
"step": 3300
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 0.14402546474331457,
|
| 1176 |
+
"grad_norm": 110.59537506103516,
|
| 1177 |
+
"learning_rate": 4.999745019059153e-05,
|
| 1178 |
+
"loss": 532.930029296875,
|
| 1179 |
+
"step": 3320
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 0.14489308802490078,
|
| 1183 |
+
"grad_norm": 140.21958923339844,
|
| 1184 |
+
"learning_rate": 4.999704605650797e-05,
|
| 1185 |
+
"loss": 528.147265625,
|
| 1186 |
+
"step": 3340
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 0.14576071130648702,
|
| 1190 |
+
"grad_norm": 115.93759155273438,
|
| 1191 |
+
"learning_rate": 4.999661220910682e-05,
|
| 1192 |
+
"loss": 525.59267578125,
|
| 1193 |
+
"step": 3360
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 0.14662833458807326,
|
| 1197 |
+
"grad_norm": 127.13153076171875,
|
| 1198 |
+
"learning_rate": 4.9996148648903794e-05,
|
| 1199 |
+
"loss": 533.487939453125,
|
| 1200 |
+
"step": 3380
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 0.1474959578696595,
|
| 1204 |
+
"grad_norm": 125.50680541992188,
|
| 1205 |
+
"learning_rate": 4.999565537644989e-05,
|
| 1206 |
+
"loss": 536.21201171875,
|
| 1207 |
+
"step": 3400
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 0.1483635811512457,
|
| 1211 |
+
"grad_norm": 111.38420867919922,
|
| 1212 |
+
"learning_rate": 4.9995132392331464e-05,
|
| 1213 |
+
"loss": 537.491796875,
|
| 1214 |
+
"step": 3420
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 0.14923120443283194,
|
| 1218 |
+
"grad_norm": 124.7078628540039,
|
| 1219 |
+
"learning_rate": 4.999457969717017e-05,
|
| 1220 |
+
"loss": 534.454248046875,
|
| 1221 |
+
"step": 3440
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 0.15009882771441818,
|
| 1225 |
+
"grad_norm": 112.11663818359375,
|
| 1226 |
+
"learning_rate": 4.999399729162299e-05,
|
| 1227 |
+
"loss": 533.73837890625,
|
| 1228 |
+
"step": 3460
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 0.15096645099600442,
|
| 1232 |
+
"grad_norm": 110.18434143066406,
|
| 1233 |
+
"learning_rate": 4.9993385176382204e-05,
|
| 1234 |
+
"loss": 526.227197265625,
|
| 1235 |
+
"step": 3480
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 0.15183407427759066,
|
| 1239 |
+
"grad_norm": 119.18470001220703,
|
| 1240 |
+
"learning_rate": 4.9992743352175406e-05,
|
| 1241 |
+
"loss": 531.402294921875,
|
| 1242 |
+
"step": 3500
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 0.15270169755917687,
|
| 1246 |
+
"grad_norm": 108.84284210205078,
|
| 1247 |
+
"learning_rate": 4.999207181976553e-05,
|
| 1248 |
+
"loss": 525.888916015625,
|
| 1249 |
+
"step": 3520
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 0.1535693208407631,
|
| 1253 |
+
"grad_norm": 138.86341857910156,
|
| 1254 |
+
"learning_rate": 4.9991370579950816e-05,
|
| 1255 |
+
"loss": 528.5728515625,
|
| 1256 |
+
"step": 3540
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 0.15443694412234935,
|
| 1260 |
+
"grad_norm": 126.5429916381836,
|
| 1261 |
+
"learning_rate": 4.999063963356479e-05,
|
| 1262 |
+
"loss": 522.498193359375,
|
| 1263 |
+
"step": 3560
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 0.15530456740393558,
|
| 1267 |
+
"grad_norm": 110.31175231933594,
|
| 1268 |
+
"learning_rate": 4.998987898147632e-05,
|
| 1269 |
+
"loss": 518.47783203125,
|
| 1270 |
+
"step": 3580
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 0.1561721906855218,
|
| 1274 |
+
"grad_norm": 128.57266235351562,
|
| 1275 |
+
"learning_rate": 4.998908862458957e-05,
|
| 1276 |
+
"loss": 525.023291015625,
|
| 1277 |
+
"step": 3600
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 0.15703981396710803,
|
| 1281 |
+
"grad_norm": 130.7259063720703,
|
| 1282 |
+
"learning_rate": 4.998826856384402e-05,
|
| 1283 |
+
"loss": 525.17392578125,
|
| 1284 |
+
"step": 3620
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 0.15790743724869427,
|
| 1288 |
+
"grad_norm": 115.61829376220703,
|
| 1289 |
+
"learning_rate": 4.998741880021444e-05,
|
| 1290 |
+
"loss": 524.6423828125,
|
| 1291 |
+
"step": 3640
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 0.1587750605302805,
|
| 1295 |
+
"grad_norm": 146.49696350097656,
|
| 1296 |
+
"learning_rate": 4.998653933471094e-05,
|
| 1297 |
+
"loss": 526.362353515625,
|
| 1298 |
+
"step": 3660
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 0.15964268381186675,
|
| 1302 |
+
"grad_norm": 113.15586853027344,
|
| 1303 |
+
"learning_rate": 4.998563016837891e-05,
|
| 1304 |
+
"loss": 516.53994140625,
|
| 1305 |
+
"step": 3680
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 0.16051030709345296,
|
| 1309 |
+
"grad_norm": 131.8157501220703,
|
| 1310 |
+
"learning_rate": 4.998469130229904e-05,
|
| 1311 |
+
"loss": 522.711962890625,
|
| 1312 |
+
"step": 3700
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 0.22592908210616616,
|
| 1316 |
+
"grad_norm": 106.28179931640625,
|
| 1317 |
+
"learning_rate": 4.999999473993204e-05,
|
| 1318 |
+
"loss": 509.725,
|
| 1319 |
+
"step": 3720
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 0.22714375459060793,
|
| 1323 |
+
"grad_norm": 103.63924407958984,
|
| 1324 |
+
"learning_rate": 4.9999975082933945e-05,
|
| 1325 |
+
"loss": 497.84814453125,
|
| 1326 |
+
"step": 3740
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 0.2283584270750497,
|
| 1330 |
+
"grad_norm": 115.9745864868164,
|
| 1331 |
+
"learning_rate": 4.99999408652088e-05,
|
| 1332 |
+
"loss": 507.36953125,
|
| 1333 |
+
"step": 3760
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 0.22957309955949143,
|
| 1337 |
+
"grad_norm": 135.81298828125,
|
| 1338 |
+
"learning_rate": 4.999989208677655e-05,
|
| 1339 |
+
"loss": 494.7404296875,
|
| 1340 |
+
"step": 3780
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 0.2307877720439332,
|
| 1344 |
+
"grad_norm": 100.2100830078125,
|
| 1345 |
+
"learning_rate": 4.999982874766558e-05,
|
| 1346 |
+
"loss": 506.553466796875,
|
| 1347 |
+
"step": 3800
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 0.23200244452837493,
|
| 1351 |
+
"grad_norm": 104.41744232177734,
|
| 1352 |
+
"learning_rate": 4.999975084791281e-05,
|
| 1353 |
+
"loss": 500.2984375,
|
| 1354 |
+
"step": 3820
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 0.2332171170128167,
|
| 1358 |
+
"grad_norm": 121.0761947631836,
|
| 1359 |
+
"learning_rate": 4.9999658387563584e-05,
|
| 1360 |
+
"loss": 505.133544921875,
|
| 1361 |
+
"step": 3840
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 0.23443178949725846,
|
| 1365 |
+
"grad_norm": 108.8710708618164,
|
| 1366 |
+
"learning_rate": 4.999955706352832e-05,
|
| 1367 |
+
"loss": 500.802880859375,
|
| 1368 |
+
"step": 3860
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 0.2356464619817002,
|
| 1372 |
+
"grad_norm": 100.03910064697266,
|
| 1373 |
+
"learning_rate": 4.999943621017866e-05,
|
| 1374 |
+
"loss": 494.677685546875,
|
| 1375 |
+
"step": 3880
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 0.23686113446614196,
|
| 1379 |
+
"grad_norm": 106.42415618896484,
|
| 1380 |
+
"learning_rate": 4.9999300796415804e-05,
|
| 1381 |
+
"loss": 495.926708984375,
|
| 1382 |
+
"step": 3900
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 0.2380758069505837,
|
| 1386 |
+
"grad_norm": 104.84850311279297,
|
| 1387 |
+
"learning_rate": 4.999915082231863e-05,
|
| 1388 |
+
"loss": 497.875634765625,
|
| 1389 |
+
"step": 3920
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 0.23929047943502546,
|
| 1393 |
+
"grad_norm": 110.93647766113281,
|
| 1394 |
+
"learning_rate": 4.9998986287974494e-05,
|
| 1395 |
+
"loss": 496.6908203125,
|
| 1396 |
+
"step": 3940
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 0.24050515191946722,
|
| 1400 |
+
"grad_norm": 112.61176300048828,
|
| 1401 |
+
"learning_rate": 4.999880719347921e-05,
|
| 1402 |
+
"loss": 495.96689453125,
|
| 1403 |
+
"step": 3960
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 0.24171982440390896,
|
| 1407 |
+
"grad_norm": 152.41073608398438,
|
| 1408 |
+
"learning_rate": 4.9998613538937095e-05,
|
| 1409 |
+
"loss": 492.7263671875,
|
| 1410 |
+
"step": 3980
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 0.24293449688835073,
|
| 1414 |
+
"grad_norm": 114.2176284790039,
|
| 1415 |
+
"learning_rate": 4.999840532446094e-05,
|
| 1416 |
+
"loss": 483.372607421875,
|
| 1417 |
+
"step": 4000
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 0.24414916937279246,
|
| 1421 |
+
"grad_norm": 97.44127655029297,
|
| 1422 |
+
"learning_rate": 4.999818255017202e-05,
|
| 1423 |
+
"loss": 492.37529296875,
|
| 1424 |
+
"step": 4020
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 0.24536384185723423,
|
| 1428 |
+
"grad_norm": 124.07328033447266,
|
| 1429 |
+
"learning_rate": 4.999794521620008e-05,
|
| 1430 |
+
"loss": 485.87412109375,
|
| 1431 |
+
"step": 4040
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 0.246578514341676,
|
| 1435 |
+
"grad_norm": 106.4791030883789,
|
| 1436 |
+
"learning_rate": 4.9997693322683346e-05,
|
| 1437 |
+
"loss": 491.435498046875,
|
| 1438 |
+
"step": 4060
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 0.24779318682611773,
|
| 1442 |
+
"grad_norm": 117.6468505859375,
|
| 1443 |
+
"learning_rate": 4.999742686976853e-05,
|
| 1444 |
+
"loss": 482.93173828125,
|
| 1445 |
+
"step": 4080
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 0.2490078593105595,
|
| 1449 |
+
"grad_norm": 117.31262969970703,
|
| 1450 |
+
"learning_rate": 4.9997145857610836e-05,
|
| 1451 |
+
"loss": 484.386865234375,
|
| 1452 |
+
"step": 4100
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 0.25022253179500126,
|
| 1456 |
+
"grad_norm": 131.6193084716797,
|
| 1457 |
+
"learning_rate": 4.999685028637392e-05,
|
| 1458 |
+
"loss": 487.89892578125,
|
| 1459 |
+
"step": 4120
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 0.251437204279443,
|
| 1463 |
+
"grad_norm": 123.34525299072266,
|
| 1464 |
+
"learning_rate": 4.999654015622993e-05,
|
| 1465 |
+
"loss": 488.15224609375,
|
| 1466 |
+
"step": 4140
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 0.25265187676388473,
|
| 1470 |
+
"grad_norm": 111.39379119873047,
|
| 1471 |
+
"learning_rate": 4.9996215467359504e-05,
|
| 1472 |
+
"loss": 481.15263671875,
|
| 1473 |
+
"step": 4160
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 0.2538665492483265,
|
| 1477 |
+
"grad_norm": 110.37937927246094,
|
| 1478 |
+
"learning_rate": 4.9995876219951744e-05,
|
| 1479 |
+
"loss": 484.3240234375,
|
| 1480 |
+
"step": 4180
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 0.25508122173276826,
|
| 1484 |
+
"grad_norm": 126.56339263916016,
|
| 1485 |
+
"learning_rate": 4.999552241420424e-05,
|
| 1486 |
+
"loss": 482.1533203125,
|
| 1487 |
+
"step": 4200
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 0.25629589421721,
|
| 1491 |
+
"grad_norm": 100.58653259277344,
|
| 1492 |
+
"learning_rate": 4.9995154050323054e-05,
|
| 1493 |
+
"loss": 479.80693359375,
|
| 1494 |
+
"step": 4220
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 0.2575105667016518,
|
| 1498 |
+
"grad_norm": 109.23738861083984,
|
| 1499 |
+
"learning_rate": 4.999477112852275e-05,
|
| 1500 |
+
"loss": 476.910205078125,
|
| 1501 |
+
"step": 4240
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 0.2587252391860935,
|
| 1505 |
+
"grad_norm": 120.3822250366211,
|
| 1506 |
+
"learning_rate": 4.999437364902633e-05,
|
| 1507 |
+
"loss": 479.670556640625,
|
| 1508 |
+
"step": 4260
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 0.25993991167053526,
|
| 1512 |
+
"grad_norm": 106.20853424072266,
|
| 1513 |
+
"learning_rate": 4.999396161206531e-05,
|
| 1514 |
+
"loss": 478.95908203125,
|
| 1515 |
+
"step": 4280
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 0.261154584154977,
|
| 1519 |
+
"grad_norm": 99.83069610595703,
|
| 1520 |
+
"learning_rate": 4.9993535017879663e-05,
|
| 1521 |
+
"loss": 481.247998046875,
|
| 1522 |
+
"step": 4300
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 0.2623692566394188,
|
| 1526 |
+
"grad_norm": 108.5458984375,
|
| 1527 |
+
"learning_rate": 4.9993093866717866e-05,
|
| 1528 |
+
"loss": 478.165966796875,
|
| 1529 |
+
"step": 4320
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 0.2635839291238605,
|
| 1533 |
+
"grad_norm": 107.90129852294922,
|
| 1534 |
+
"learning_rate": 4.999263815883685e-05,
|
| 1535 |
+
"loss": 486.360400390625,
|
| 1536 |
+
"step": 4340
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 0.26479860160830226,
|
| 1540 |
+
"grad_norm": 116.5055160522461,
|
| 1541 |
+
"learning_rate": 4.999216789450203e-05,
|
| 1542 |
+
"loss": 471.4162109375,
|
| 1543 |
+
"step": 4360
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 0.26601327409274406,
|
| 1547 |
+
"grad_norm": 131.4321746826172,
|
| 1548 |
+
"learning_rate": 4.99916830739873e-05,
|
| 1549 |
+
"loss": 469.025537109375,
|
| 1550 |
+
"step": 4380
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 0.2672279465771858,
|
| 1554 |
+
"grad_norm": 118.11991882324219,
|
| 1555 |
+
"learning_rate": 4.999118369757505e-05,
|
| 1556 |
+
"loss": 483.506201171875,
|
| 1557 |
+
"step": 4400
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 0.26844261906162753,
|
| 1561 |
+
"grad_norm": 106.02652740478516,
|
| 1562 |
+
"learning_rate": 4.999066976555611e-05,
|
| 1563 |
+
"loss": 478.2771484375,
|
| 1564 |
+
"step": 4420
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 0.2696572915460693,
|
| 1568 |
+
"grad_norm": 117.4029541015625,
|
| 1569 |
+
"learning_rate": 4.9990141278229825e-05,
|
| 1570 |
+
"loss": 477.311279296875,
|
| 1571 |
+
"step": 4440
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 0.27087196403051106,
|
| 1575 |
+
"grad_norm": 132.9343719482422,
|
| 1576 |
+
"learning_rate": 4.998959823590399e-05,
|
| 1577 |
+
"loss": 472.3001953125,
|
| 1578 |
+
"step": 4460
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 0.2720866365149528,
|
| 1582 |
+
"grad_norm": 125.26654052734375,
|
| 1583 |
+
"learning_rate": 4.9989040638894904e-05,
|
| 1584 |
+
"loss": 473.738671875,
|
| 1585 |
+
"step": 4480
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 0.2733013089993946,
|
| 1589 |
+
"grad_norm": 130.5485076904297,
|
| 1590 |
+
"learning_rate": 4.9988468487527306e-05,
|
| 1591 |
+
"loss": 468.090087890625,
|
| 1592 |
+
"step": 4500
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 0.2745159814838363,
|
| 1596 |
+
"grad_norm": 99.37794494628906,
|
| 1597 |
+
"learning_rate": 4.998788178213446e-05,
|
| 1598 |
+
"loss": 476.109765625,
|
| 1599 |
+
"step": 4520
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 0.27573065396827806,
|
| 1603 |
+
"grad_norm": 108.37550354003906,
|
| 1604 |
+
"learning_rate": 4.998728052305806e-05,
|
| 1605 |
+
"loss": 473.056884765625,
|
| 1606 |
+
"step": 4540
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 0.2769453264527198,
|
| 1610 |
+
"grad_norm": 107.37875366210938,
|
| 1611 |
+
"learning_rate": 4.99866647106483e-05,
|
| 1612 |
+
"loss": 479.7654296875,
|
| 1613 |
+
"step": 4560
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 0.2781599989371616,
|
| 1617 |
+
"grad_norm": 134.2859649658203,
|
| 1618 |
+
"learning_rate": 4.998603434526387e-05,
|
| 1619 |
+
"loss": 473.32333984375,
|
| 1620 |
+
"step": 4580
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 0.2793746714216033,
|
| 1624 |
+
"grad_norm": 130.39781188964844,
|
| 1625 |
+
"learning_rate": 4.998538942727188e-05,
|
| 1626 |
+
"loss": 473.8544921875,
|
| 1627 |
+
"step": 4600
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 0.2922806590232955,
|
| 1631 |
+
"grad_norm": 82.23613739013672,
|
| 1632 |
+
"learning_rate": 4.999999802498301e-05,
|
| 1633 |
+
"loss": 224.93673001802884,
|
| 1634 |
+
"step": 4620
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 0.28180401639048686,
|
| 1638 |
+
"grad_norm": Infinity,
|
| 1639 |
+
"learning_rate": 4.999999998736047e-05,
|
| 1640 |
+
"loss": 522.9669189453125,
|
| 1641 |
+
"step": 4640
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 0.2830186888749286,
|
| 1645 |
+
"grad_norm": 128.01324462890625,
|
| 1646 |
+
"learning_rate": 4.9999995437128654e-05,
|
| 1647 |
+
"loss": 509.814208984375,
|
| 1648 |
+
"step": 4660
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 0.28423336135937033,
|
| 1652 |
+
"grad_norm": 124.84363555908203,
|
| 1653 |
+
"learning_rate": 4.9999980775272446e-05,
|
| 1654 |
+
"loss": 506.356494140625,
|
| 1655 |
+
"step": 4680
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 0.2854480338438121,
|
| 1659 |
+
"grad_norm": 147.43655395507812,
|
| 1660 |
+
"learning_rate": 4.9999956001797785e-05,
|
| 1661 |
+
"loss": 513.00703125,
|
| 1662 |
+
"step": 4700
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 0.28666270632825386,
|
| 1666 |
+
"grad_norm": 146.66969299316406,
|
| 1667 |
+
"learning_rate": 4.9999921116714695e-05,
|
| 1668 |
+
"loss": 507.043310546875,
|
| 1669 |
+
"step": 4720
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 0.2878773788126956,
|
| 1673 |
+
"grad_norm": 107.6183853149414,
|
| 1674 |
+
"learning_rate": 4.9999876120037284e-05,
|
| 1675 |
+
"loss": 506.249169921875,
|
| 1676 |
+
"step": 4740
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 0.28909205129713733,
|
| 1680 |
+
"grad_norm": 135.1703338623047,
|
| 1681 |
+
"learning_rate": 4.999982101178373e-05,
|
| 1682 |
+
"loss": 498.55283203125,
|
| 1683 |
+
"step": 4760
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 0.2903067237815791,
|
| 1687 |
+
"grad_norm": 129.0010986328125,
|
| 1688 |
+
"learning_rate": 4.999975579197635e-05,
|
| 1689 |
+
"loss": 494.26435546875,
|
| 1690 |
+
"step": 4780
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 0.29152139626602086,
|
| 1694 |
+
"grad_norm": 104.45521545410156,
|
| 1695 |
+
"learning_rate": 4.999968046064152e-05,
|
| 1696 |
+
"loss": 500.8259765625,
|
| 1697 |
+
"step": 4800
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 0.2927360687504626,
|
| 1701 |
+
"grad_norm": 122.8540267944336,
|
| 1702 |
+
"learning_rate": 4.9999595017809696e-05,
|
| 1703 |
+
"loss": 491.134375,
|
| 1704 |
+
"step": 4820
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 0.2939507412349044,
|
| 1708 |
+
"grad_norm": 110.6492919921875,
|
| 1709 |
+
"learning_rate": 4.999949946351544e-05,
|
| 1710 |
+
"loss": 491.937548828125,
|
| 1711 |
+
"step": 4840
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 0.2951654137193461,
|
| 1715 |
+
"grad_norm": 122.96728515625,
|
| 1716 |
+
"learning_rate": 4.9999393797797404e-05,
|
| 1717 |
+
"loss": 493.3181640625,
|
| 1718 |
+
"step": 4860
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 0.29638008620378786,
|
| 1722 |
+
"grad_norm": 109.75753784179688,
|
| 1723 |
+
"learning_rate": 4.999927802069833e-05,
|
| 1724 |
+
"loss": 495.562451171875,
|
| 1725 |
+
"step": 4880
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 0.29759475868822965,
|
| 1729 |
+
"grad_norm": 113.73827362060547,
|
| 1730 |
+
"learning_rate": 4.999915213226504e-05,
|
| 1731 |
+
"loss": 494.584033203125,
|
| 1732 |
+
"step": 4900
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 0.2988094311726714,
|
| 1736 |
+
"grad_norm": 166.45193481445312,
|
| 1737 |
+
"learning_rate": 4.999901613254845e-05,
|
| 1738 |
+
"loss": 497.014306640625,
|
| 1739 |
+
"step": 4920
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 0.30002410365711313,
|
| 1743 |
+
"grad_norm": 125.07030487060547,
|
| 1744 |
+
"learning_rate": 4.999887002160356e-05,
|
| 1745 |
+
"loss": 491.55986328125,
|
| 1746 |
+
"step": 4940
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 0.3012387761415549,
|
| 1750 |
+
"grad_norm": 101.7028579711914,
|
| 1751 |
+
"learning_rate": 4.999871379948949e-05,
|
| 1752 |
+
"loss": 495.484912109375,
|
| 1753 |
+
"step": 4960
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 0.30245344862599666,
|
| 1757 |
+
"grad_norm": 120.32843017578125,
|
| 1758 |
+
"learning_rate": 4.9998547466269406e-05,
|
| 1759 |
+
"loss": 488.6263671875,
|
| 1760 |
+
"step": 4980
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 0.3036681211104384,
|
| 1764 |
+
"grad_norm": 117.14525604248047,
|
| 1765 |
+
"learning_rate": 4.999837102201059e-05,
|
| 1766 |
+
"loss": 488.448876953125,
|
| 1767 |
+
"step": 5000
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 0.30488279359488013,
|
| 1771 |
+
"grad_norm": 129.26194763183594,
|
| 1772 |
+
"learning_rate": 4.9998184466784416e-05,
|
| 1773 |
+
"loss": 489.488525390625,
|
| 1774 |
+
"step": 5020
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 0.3060974660793219,
|
| 1778 |
+
"grad_norm": 148.86119079589844,
|
| 1779 |
+
"learning_rate": 4.9997987800666326e-05,
|
| 1780 |
+
"loss": 488.17734375,
|
| 1781 |
+
"step": 5040
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 0.30731213856376366,
|
| 1785 |
+
"grad_norm": 125.63531494140625,
|
| 1786 |
+
"learning_rate": 4.999778102373587e-05,
|
| 1787 |
+
"loss": 482.14873046875,
|
| 1788 |
+
"step": 5060
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 0.3085268110482054,
|
| 1792 |
+
"grad_norm": 128.2238311767578,
|
| 1793 |
+
"learning_rate": 4.999756413607668e-05,
|
| 1794 |
+
"loss": 495.12177734375,
|
| 1795 |
+
"step": 5080
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 0.3097414835326472,
|
| 1799 |
+
"grad_norm": 139.5421905517578,
|
| 1800 |
+
"learning_rate": 4.999733713777648e-05,
|
| 1801 |
+
"loss": 488.80283203125,
|
| 1802 |
+
"step": 5100
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 0.3109561560170889,
|
| 1806 |
+
"grad_norm": 118.46855163574219,
|
| 1807 |
+
"learning_rate": 4.9997100028927086e-05,
|
| 1808 |
+
"loss": 492.863134765625,
|
| 1809 |
+
"step": 5120
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 0.31217082850153066,
|
| 1813 |
+
"grad_norm": 109.22224426269531,
|
| 1814 |
+
"learning_rate": 4.999685280962439e-05,
|
| 1815 |
+
"loss": 493.7431640625,
|
| 1816 |
+
"step": 5140
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 0.31338550098597245,
|
| 1820 |
+
"grad_norm": 114.17903900146484,
|
| 1821 |
+
"learning_rate": 4.999659547996841e-05,
|
| 1822 |
+
"loss": 485.016796875,
|
| 1823 |
+
"step": 5160
|
| 1824 |
+
},
|
| 1825 |
+
{
|
| 1826 |
+
"epoch": 0.3146001734704142,
|
| 1827 |
+
"grad_norm": 109.17562103271484,
|
| 1828 |
+
"learning_rate": 4.99963280400632e-05,
|
| 1829 |
+
"loss": 488.472265625,
|
| 1830 |
+
"step": 5180
|
| 1831 |
+
},
|
| 1832 |
+
{
|
| 1833 |
+
"epoch": 0.3158148459548559,
|
| 1834 |
+
"grad_norm": 115.13211059570312,
|
| 1835 |
+
"learning_rate": 4.999605049001694e-05,
|
| 1836 |
+
"loss": 479.812939453125,
|
| 1837 |
+
"step": 5200
|
| 1838 |
+
},
|
| 1839 |
+
{
|
| 1840 |
+
"epoch": 0.31702951843929766,
|
| 1841 |
+
"grad_norm": 139.9939727783203,
|
| 1842 |
+
"learning_rate": 4.9995762829941885e-05,
|
| 1843 |
+
"loss": 485.855517578125,
|
| 1844 |
+
"step": 5220
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 0.31824419092373946,
|
| 1848 |
+
"grad_norm": 135.5021514892578,
|
| 1849 |
+
"learning_rate": 4.999546505995439e-05,
|
| 1850 |
+
"loss": 488.666552734375,
|
| 1851 |
+
"step": 5240
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 0.3194588634081812,
|
| 1855 |
+
"grad_norm": 99.49887084960938,
|
| 1856 |
+
"learning_rate": 4.9995157180174886e-05,
|
| 1857 |
+
"loss": 471.69619140625,
|
| 1858 |
+
"step": 5260
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 0.32067353589262293,
|
| 1862 |
+
"grad_norm": 110.44123077392578,
|
| 1863 |
+
"learning_rate": 4.9994839190727904e-05,
|
| 1864 |
+
"loss": 480.227587890625,
|
| 1865 |
+
"step": 5280
|
| 1866 |
+
},
|
| 1867 |
+
{
|
| 1868 |
+
"epoch": 0.3218882083770647,
|
| 1869 |
+
"grad_norm": 117.12146759033203,
|
| 1870 |
+
"learning_rate": 4.999451109174206e-05,
|
| 1871 |
+
"loss": 484.912548828125,
|
| 1872 |
+
"step": 5300
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 0.32310288086150646,
|
| 1876 |
+
"grad_norm": 104.73037719726562,
|
| 1877 |
+
"learning_rate": 4.9994172883350054e-05,
|
| 1878 |
+
"loss": 480.243798828125,
|
| 1879 |
+
"step": 5320
|
| 1880 |
+
},
|
| 1881 |
+
{
|
| 1882 |
+
"epoch": 0.3243175533459482,
|
| 1883 |
+
"grad_norm": 111.67479705810547,
|
| 1884 |
+
"learning_rate": 4.9993824565688685e-05,
|
| 1885 |
+
"loss": 482.366259765625,
|
| 1886 |
+
"step": 5340
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 0.32553222583039,
|
| 1890 |
+
"grad_norm": 137.9676513671875,
|
| 1891 |
+
"learning_rate": 4.999346613889883e-05,
|
| 1892 |
+
"loss": 487.56142578125,
|
| 1893 |
+
"step": 5360
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 0.3267468983148317,
|
| 1897 |
+
"grad_norm": 125.0929183959961,
|
| 1898 |
+
"learning_rate": 4.9993097603125466e-05,
|
| 1899 |
+
"loss": 480.0265625,
|
| 1900 |
+
"step": 5380
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"epoch": 0.32796157079927346,
|
| 1904 |
+
"grad_norm": 133.99122619628906,
|
| 1905 |
+
"learning_rate": 4.999271895851765e-05,
|
| 1906 |
+
"loss": 473.62099609375,
|
| 1907 |
+
"step": 5400
|
| 1908 |
+
},
|
| 1909 |
+
{
|
| 1910 |
+
"epoch": 0.32917624328371525,
|
| 1911 |
+
"grad_norm": 128.70616149902344,
|
| 1912 |
+
"learning_rate": 4.9992330205228536e-05,
|
| 1913 |
+
"loss": 476.02734375,
|
| 1914 |
+
"step": 5420
|
| 1915 |
+
},
|
| 1916 |
+
{
|
| 1917 |
+
"epoch": 0.330390915768157,
|
| 1918 |
+
"grad_norm": 143.84982299804688,
|
| 1919 |
+
"learning_rate": 4.999193134341534e-05,
|
| 1920 |
+
"loss": 467.562109375,
|
| 1921 |
+
"step": 5440
|
| 1922 |
+
},
|
| 1923 |
+
{
|
| 1924 |
+
"epoch": 0.3316055882525987,
|
| 1925 |
+
"grad_norm": 168.56112670898438,
|
| 1926 |
+
"learning_rate": 4.9991522373239414e-05,
|
| 1927 |
+
"loss": 477.2583984375,
|
| 1928 |
+
"step": 5460
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 0.33282026073704046,
|
| 1932 |
+
"grad_norm": 136.1503143310547,
|
| 1933 |
+
"learning_rate": 4.999110329486616e-05,
|
| 1934 |
+
"loss": 474.83203125,
|
| 1935 |
+
"step": 5480
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 0.33403493322148226,
|
| 1939 |
+
"grad_norm": 116.25304412841797,
|
| 1940 |
+
"learning_rate": 4.9990674108465074e-05,
|
| 1941 |
+
"loss": 476.651708984375,
|
| 1942 |
+
"step": 5500
|
| 1943 |
+
},
|
| 1944 |
+
{
|
| 1945 |
+
"epoch": 0.335249605705924,
|
| 1946 |
+
"grad_norm": 138.05633544921875,
|
| 1947 |
+
"learning_rate": 4.999023481420977e-05,
|
| 1948 |
+
"loss": 472.470263671875,
|
| 1949 |
+
"step": 5520
|
| 1950 |
+
},
|
| 1951 |
+
{
|
| 1952 |
+
"epoch": 0.33646427819036573,
|
| 1953 |
+
"grad_norm": 106.01046752929688,
|
| 1954 |
+
"learning_rate": 4.99897854122779e-05,
|
| 1955 |
+
"loss": 472.5498046875,
|
| 1956 |
+
"step": 5540
|
| 1957 |
+
},
|
| 1958 |
+
{
|
| 1959 |
+
"epoch": 0.3376789506748075,
|
| 1960 |
+
"grad_norm": 116.29874420166016,
|
| 1961 |
+
"learning_rate": 4.998932590285125e-05,
|
| 1962 |
+
"loss": 471.28046875,
|
| 1963 |
+
"step": 5560
|
| 1964 |
+
},
|
| 1965 |
+
{
|
| 1966 |
+
"epoch": 0.33889362315924926,
|
| 1967 |
+
"grad_norm": 108.59446716308594,
|
| 1968 |
+
"learning_rate": 4.998885628611566e-05,
|
| 1969 |
+
"loss": 473.924267578125,
|
| 1970 |
+
"step": 5580
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 0.340108295643691,
|
| 1974 |
+
"grad_norm": 108.22511291503906,
|
| 1975 |
+
"learning_rate": 4.9988376562261085e-05,
|
| 1976 |
+
"loss": 469.3677734375,
|
| 1977 |
+
"step": 5600
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 0.3413229681281328,
|
| 1981 |
+
"grad_norm": 119.11192321777344,
|
| 1982 |
+
"learning_rate": 4.9999999949441866e-05,
|
| 1983 |
+
"loss": 457.769189453125,
|
| 1984 |
+
"step": 5620
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 0.3425376406125745,
|
| 1988 |
+
"grad_norm": 114.36215209960938,
|
| 1989 |
+
"learning_rate": 4.9999993882466175e-05,
|
| 1990 |
+
"loss": 459.95791015625,
|
| 1991 |
+
"step": 5640
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"epoch": 0.34375231309701626,
|
| 1995 |
+
"grad_norm": 121.7007827758789,
|
| 1996 |
+
"learning_rate": 4.9999977703866726e-05,
|
| 1997 |
+
"loss": 457.78818359375,
|
| 1998 |
+
"step": 5660
|
| 1999 |
+
},
|
| 2000 |
+
{
|
| 2001 |
+
"epoch": 0.34496698558145805,
|
| 2002 |
+
"grad_norm": 125.85967254638672,
|
| 2003 |
+
"learning_rate": 4.9999951413650066e-05,
|
| 2004 |
+
"loss": 459.223193359375,
|
| 2005 |
+
"step": 5680
|
| 2006 |
+
},
|
| 2007 |
+
{
|
| 2008 |
+
"epoch": 0.3461816580658998,
|
| 2009 |
+
"grad_norm": 117.03638458251953,
|
| 2010 |
+
"learning_rate": 4.9999915011826826e-05,
|
| 2011 |
+
"loss": 452.325439453125,
|
| 2012 |
+
"step": 5700
|
| 2013 |
+
},
|
| 2014 |
+
{
|
| 2015 |
+
"epoch": 0.3473963305503415,
|
| 2016 |
+
"grad_norm": 111.20164489746094,
|
| 2017 |
+
"learning_rate": 4.999986849841173e-05,
|
| 2018 |
+
"loss": 456.363916015625,
|
| 2019 |
+
"step": 5720
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 0.34861100303478326,
|
| 2023 |
+
"grad_norm": 106.04707336425781,
|
| 2024 |
+
"learning_rate": 4.9999811873423586e-05,
|
| 2025 |
+
"loss": 449.553466796875,
|
| 2026 |
+
"step": 5740
|
| 2027 |
+
},
|
| 2028 |
+
{
|
| 2029 |
+
"epoch": 0.34982567551922505,
|
| 2030 |
+
"grad_norm": 121.10123443603516,
|
| 2031 |
+
"learning_rate": 4.9999745136885306e-05,
|
| 2032 |
+
"loss": 453.322314453125,
|
| 2033 |
+
"step": 5760
|
| 2034 |
+
},
|
| 2035 |
+
{
|
| 2036 |
+
"epoch": 0.3510403480036668,
|
| 2037 |
+
"grad_norm": 102.19346618652344,
|
| 2038 |
+
"learning_rate": 4.9999672371375163e-05,
|
| 2039 |
+
"loss": 445.2501953125,
|
| 2040 |
+
"step": 5780
|
| 2041 |
+
},
|
| 2042 |
+
{
|
| 2043 |
+
"epoch": 0.35225502048810853,
|
| 2044 |
+
"grad_norm": 126.20061492919922,
|
| 2045 |
+
"learning_rate": 4.999958591739547e-05,
|
| 2046 |
+
"loss": 451.185205078125,
|
| 2047 |
+
"step": 5800
|
| 2048 |
+
},
|
| 2049 |
+
{
|
| 2050 |
+
"epoch": 0.3534696929725503,
|
| 2051 |
+
"grad_norm": 126.69395446777344,
|
| 2052 |
+
"learning_rate": 4.9999489351957034e-05,
|
| 2053 |
+
"loss": 443.0,
|
| 2054 |
+
"step": 5820
|
| 2055 |
+
},
|
| 2056 |
+
{
|
| 2057 |
+
"epoch": 0.35468436545699206,
|
| 2058 |
+
"grad_norm": 120.10934448242188,
|
| 2059 |
+
"learning_rate": 4.9999382675098905e-05,
|
| 2060 |
+
"loss": 439.477001953125,
|
| 2061 |
+
"step": 5840
|
| 2062 |
+
},
|
| 2063 |
+
{
|
| 2064 |
+
"epoch": 0.3558990379414338,
|
| 2065 |
+
"grad_norm": 109.93563079833984,
|
| 2066 |
+
"learning_rate": 4.999926588686423e-05,
|
| 2067 |
+
"loss": 439.631103515625,
|
| 2068 |
+
"step": 5860
|
| 2069 |
+
},
|
| 2070 |
+
{
|
| 2071 |
+
"epoch": 0.3571137104258756,
|
| 2072 |
+
"grad_norm": 117.47706604003906,
|
| 2073 |
+
"learning_rate": 4.9999138987300246e-05,
|
| 2074 |
+
"loss": 441.124560546875,
|
| 2075 |
+
"step": 5880
|
| 2076 |
+
},
|
| 2077 |
+
{
|
| 2078 |
+
"epoch": 0.3583283829103173,
|
| 2079 |
+
"grad_norm": 99.2137222290039,
|
| 2080 |
+
"learning_rate": 4.999900197645829e-05,
|
| 2081 |
+
"loss": 439.1115234375,
|
| 2082 |
+
"step": 5900
|
| 2083 |
+
},
|
| 2084 |
+
{
|
| 2085 |
+
"epoch": 0.35954305539475906,
|
| 2086 |
+
"grad_norm": 97.20148468017578,
|
| 2087 |
+
"learning_rate": 4.999885485439376e-05,
|
| 2088 |
+
"loss": 446.5513671875,
|
| 2089 |
+
"step": 5920
|
| 2090 |
+
},
|
| 2091 |
+
{
|
| 2092 |
+
"epoch": 0.3607577278792008,
|
| 2093 |
+
"grad_norm": 133.8880615234375,
|
| 2094 |
+
"learning_rate": 4.999869762116617e-05,
|
| 2095 |
+
"loss": 436.385302734375,
|
| 2096 |
+
"step": 5940
|
| 2097 |
+
},
|
| 2098 |
+
{
|
| 2099 |
+
"epoch": 0.3619724003636426,
|
| 2100 |
+
"grad_norm": 110.88639831542969,
|
| 2101 |
+
"learning_rate": 4.999853027683912e-05,
|
| 2102 |
+
"loss": 438.487548828125,
|
| 2103 |
+
"step": 5960
|
| 2104 |
+
},
|
| 2105 |
+
{
|
| 2106 |
+
"epoch": 0.3631870728480843,
|
| 2107 |
+
"grad_norm": 122.55818939208984,
|
| 2108 |
+
"learning_rate": 4.9998352821480295e-05,
|
| 2109 |
+
"loss": 433.748681640625,
|
| 2110 |
+
"step": 5980
|
| 2111 |
+
},
|
| 2112 |
+
{
|
| 2113 |
+
"epoch": 0.36440174533252606,
|
| 2114 |
+
"grad_norm": 112.45657348632812,
|
| 2115 |
+
"learning_rate": 4.9998165255161455e-05,
|
| 2116 |
+
"loss": 446.138623046875,
|
| 2117 |
+
"step": 6000
|
| 2118 |
+
},
|
| 2119 |
+
{
|
| 2120 |
+
"epoch": 0.36561641781696785,
|
| 2121 |
+
"grad_norm": 112.1351089477539,
|
| 2122 |
+
"learning_rate": 4.999796757795848e-05,
|
| 2123 |
+
"loss": 441.529248046875,
|
| 2124 |
+
"step": 6020
|
| 2125 |
+
},
|
| 2126 |
+
{
|
| 2127 |
+
"epoch": 0.3668310903014096,
|
| 2128 |
+
"grad_norm": 109.1108169555664,
|
| 2129 |
+
"learning_rate": 4.999775978995133e-05,
|
| 2130 |
+
"loss": 435.71748046875,
|
| 2131 |
+
"step": 6040
|
| 2132 |
+
},
|
| 2133 |
+
{
|
| 2134 |
+
"epoch": 0.3680457627858513,
|
| 2135 |
+
"grad_norm": 113.85139465332031,
|
| 2136 |
+
"learning_rate": 4.999754189122402e-05,
|
| 2137 |
+
"loss": 441.027294921875,
|
| 2138 |
+
"step": 6060
|
| 2139 |
+
},
|
| 2140 |
+
{
|
| 2141 |
+
"epoch": 0.3692604352702931,
|
| 2142 |
+
"grad_norm": 146.5244598388672,
|
| 2143 |
+
"learning_rate": 4.99973138818647e-05,
|
| 2144 |
+
"loss": 436.488623046875,
|
| 2145 |
+
"step": 6080
|
| 2146 |
+
},
|
| 2147 |
+
{
|
| 2148 |
+
"epoch": 0.37047510775473486,
|
| 2149 |
+
"grad_norm": 103.91339111328125,
|
| 2150 |
+
"learning_rate": 4.999707576196559e-05,
|
| 2151 |
+
"loss": 440.46455078125,
|
| 2152 |
+
"step": 6100
|
| 2153 |
+
},
|
| 2154 |
+
{
|
| 2155 |
+
"epoch": 0.3716897802391766,
|
| 2156 |
+
"grad_norm": 121.91517639160156,
|
| 2157 |
+
"learning_rate": 4.9996827531623014e-05,
|
| 2158 |
+
"loss": 433.736279296875,
|
| 2159 |
+
"step": 6120
|
| 2160 |
+
},
|
| 2161 |
+
{
|
| 2162 |
+
"epoch": 0.3729044527236184,
|
| 2163 |
+
"grad_norm": 131.15545654296875,
|
| 2164 |
+
"learning_rate": 4.999656919093735e-05,
|
| 2165 |
+
"loss": 437.878955078125,
|
| 2166 |
+
"step": 6140
|
| 2167 |
+
},
|
| 2168 |
+
{
|
| 2169 |
+
"epoch": 0.3741191252080601,
|
| 2170 |
+
"grad_norm": 109.45008850097656,
|
| 2171 |
+
"learning_rate": 4.99963007400131e-05,
|
| 2172 |
+
"loss": 434.235888671875,
|
| 2173 |
+
"step": 6160
|
| 2174 |
+
},
|
| 2175 |
+
{
|
| 2176 |
+
"epoch": 0.37533379769250186,
|
| 2177 |
+
"grad_norm": 102.5931625366211,
|
| 2178 |
+
"learning_rate": 4.999602217895884e-05,
|
| 2179 |
+
"loss": 428.64306640625,
|
| 2180 |
+
"step": 6180
|
| 2181 |
+
},
|
| 2182 |
+
{
|
| 2183 |
+
"epoch": 0.3765484701769436,
|
| 2184 |
+
"grad_norm": 108.3375015258789,
|
| 2185 |
+
"learning_rate": 4.999573350788724e-05,
|
| 2186 |
+
"loss": 432.224609375,
|
| 2187 |
+
"step": 6200
|
| 2188 |
+
},
|
| 2189 |
+
{
|
| 2190 |
+
"epoch": 0.3777631426613854,
|
| 2191 |
+
"grad_norm": 105.30213928222656,
|
| 2192 |
+
"learning_rate": 4.9995434726915055e-05,
|
| 2193 |
+
"loss": 434.313720703125,
|
| 2194 |
+
"step": 6220
|
| 2195 |
+
},
|
| 2196 |
+
{
|
| 2197 |
+
"epoch": 0.3789778151458271,
|
| 2198 |
+
"grad_norm": 111.43559265136719,
|
| 2199 |
+
"learning_rate": 4.9995125836163136e-05,
|
| 2200 |
+
"loss": 426.7763671875,
|
| 2201 |
+
"step": 6240
|
| 2202 |
+
},
|
| 2203 |
+
{
|
| 2204 |
+
"epoch": 0.38019248763026886,
|
| 2205 |
+
"grad_norm": 122.2061538696289,
|
| 2206 |
+
"learning_rate": 4.999480683575641e-05,
|
| 2207 |
+
"loss": 433.124267578125,
|
| 2208 |
+
"step": 6260
|
| 2209 |
+
},
|
| 2210 |
+
{
|
| 2211 |
+
"epoch": 0.38140716011471065,
|
| 2212 |
+
"grad_norm": 128.8978729248047,
|
| 2213 |
+
"learning_rate": 4.999447772582391e-05,
|
| 2214 |
+
"loss": 430.33837890625,
|
| 2215 |
+
"step": 6280
|
| 2216 |
+
},
|
| 2217 |
+
{
|
| 2218 |
+
"epoch": 0.3826218325991524,
|
| 2219 |
+
"grad_norm": 99.80685424804688,
|
| 2220 |
+
"learning_rate": 4.999413850649875e-05,
|
| 2221 |
+
"loss": 439.00517578125,
|
| 2222 |
+
"step": 6300
|
| 2223 |
+
},
|
| 2224 |
+
{
|
| 2225 |
+
"epoch": 0.3838365050835941,
|
| 2226 |
+
"grad_norm": 128.85317993164062,
|
| 2227 |
+
"learning_rate": 4.9993789177918124e-05,
|
| 2228 |
+
"loss": 426.54765625,
|
| 2229 |
+
"step": 6320
|
| 2230 |
+
},
|
| 2231 |
+
{
|
| 2232 |
+
"epoch": 0.3850511775680359,
|
| 2233 |
+
"grad_norm": 122.83307647705078,
|
| 2234 |
+
"learning_rate": 4.999342974022333e-05,
|
| 2235 |
+
"loss": 431.87216796875,
|
| 2236 |
+
"step": 6340
|
| 2237 |
+
},
|
| 2238 |
+
{
|
| 2239 |
+
"epoch": 0.38626585005247766,
|
| 2240 |
+
"grad_norm": 110.61914825439453,
|
| 2241 |
+
"learning_rate": 4.999306019355974e-05,
|
| 2242 |
+
"loss": 428.231591796875,
|
| 2243 |
+
"step": 6360
|
| 2244 |
+
},
|
| 2245 |
+
{
|
| 2246 |
+
"epoch": 0.3874805225369194,
|
| 2247 |
+
"grad_norm": 118.25730895996094,
|
| 2248 |
+
"learning_rate": 4.999268053807683e-05,
|
| 2249 |
+
"loss": 427.494970703125,
|
| 2250 |
+
"step": 6380
|
| 2251 |
+
},
|
| 2252 |
+
{
|
| 2253 |
+
"epoch": 0.38869519502136113,
|
| 2254 |
+
"grad_norm": 159.15020751953125,
|
| 2255 |
+
"learning_rate": 4.999229077392816e-05,
|
| 2256 |
+
"loss": 428.88037109375,
|
| 2257 |
+
"step": 6400
|
| 2258 |
+
},
|
| 2259 |
+
{
|
| 2260 |
+
"epoch": 0.3899098675058029,
|
| 2261 |
+
"grad_norm": 115.5860595703125,
|
| 2262 |
+
"learning_rate": 4.999189090127138e-05,
|
| 2263 |
+
"loss": 421.6064453125,
|
| 2264 |
+
"step": 6420
|
| 2265 |
+
},
|
| 2266 |
+
{
|
| 2267 |
+
"epoch": 0.39112453999024466,
|
| 2268 |
+
"grad_norm": 109.9117202758789,
|
| 2269 |
+
"learning_rate": 4.9991480920268205e-05,
|
| 2270 |
+
"loss": 434.52431640625,
|
| 2271 |
+
"step": 6440
|
| 2272 |
+
},
|
| 2273 |
+
{
|
| 2274 |
+
"epoch": 0.3923392124746864,
|
| 2275 |
+
"grad_norm": 108.77495574951172,
|
| 2276 |
+
"learning_rate": 4.999999817990724e-05,
|
| 2277 |
+
"loss": 466.3345052083333,
|
| 2278 |
+
"step": 6460
|
| 2279 |
+
},
|
| 2280 |
+
{
|
| 2281 |
+
"epoch": 0.3935538849591282,
|
| 2282 |
+
"grad_norm": 106.6700439453125,
|
| 2283 |
+
"learning_rate": 4.9999987057119104e-05,
|
| 2284 |
+
"loss": 463.743310546875,
|
| 2285 |
+
"step": 6480
|
| 2286 |
+
},
|
| 2287 |
+
{
|
| 2288 |
+
"epoch": 0.3947685574435699,
|
| 2289 |
+
"grad_norm": 114.01268768310547,
|
| 2290 |
+
"learning_rate": 4.999996582270998e-05,
|
| 2291 |
+
"loss": 454.19453125,
|
| 2292 |
+
"step": 6500
|
| 2293 |
+
},
|
| 2294 |
+
{
|
| 2295 |
+
"epoch": 0.39598322992801166,
|
| 2296 |
+
"grad_norm": 108.73348236083984,
|
| 2297 |
+
"learning_rate": 4.999993447668844e-05,
|
| 2298 |
+
"loss": 457.237109375,
|
| 2299 |
+
"step": 6520
|
| 2300 |
+
},
|
| 2301 |
+
{
|
| 2302 |
+
"epoch": 0.39719790241245345,
|
| 2303 |
+
"grad_norm": 108.1834716796875,
|
| 2304 |
+
"learning_rate": 4.9999893019067175e-05,
|
| 2305 |
+
"loss": 453.765087890625,
|
| 2306 |
+
"step": 6540
|
| 2307 |
+
},
|
| 2308 |
+
{
|
| 2309 |
+
"epoch": 0.3984125748968952,
|
| 2310 |
+
"grad_norm": 123.11573028564453,
|
| 2311 |
+
"learning_rate": 4.999984144986295e-05,
|
| 2312 |
+
"loss": 457.87333984375,
|
| 2313 |
+
"step": 6560
|
| 2314 |
+
},
|
| 2315 |
+
{
|
| 2316 |
+
"epoch": 0.3996272473813369,
|
| 2317 |
+
"grad_norm": 111.18988800048828,
|
| 2318 |
+
"learning_rate": 4.9999779769096624e-05,
|
| 2319 |
+
"loss": 455.51279296875,
|
| 2320 |
+
"step": 6580
|
| 2321 |
+
},
|
| 2322 |
+
{
|
| 2323 |
+
"epoch": 0.4008419198657787,
|
| 2324 |
+
"grad_norm": 115.29268646240234,
|
| 2325 |
+
"learning_rate": 4.9999711806556894e-05,
|
| 2326 |
+
"loss": 447.592626953125,
|
| 2327 |
+
"step": 6600
|
| 2328 |
+
},
|
| 2329 |
+
{
|
| 2330 |
+
"epoch": 0.40205659235022045,
|
| 2331 |
+
"grad_norm": 114.5698013305664,
|
| 2332 |
+
"learning_rate": 4.999963040831994e-05,
|
| 2333 |
+
"loss": 456.120166015625,
|
| 2334 |
+
"step": 6620
|
| 2335 |
+
},
|
| 2336 |
+
{
|
| 2337 |
+
"epoch": 0.4032712648346622,
|
| 2338 |
+
"grad_norm": 144.33851623535156,
|
| 2339 |
+
"learning_rate": 4.9999538898606254e-05,
|
| 2340 |
+
"loss": 452.684326171875,
|
| 2341 |
+
"step": 6640
|
| 2342 |
+
},
|
| 2343 |
+
{
|
| 2344 |
+
"epoch": 0.4044859373191039,
|
| 2345 |
+
"grad_norm": 106.46892547607422,
|
| 2346 |
+
"learning_rate": 4.9999437277452834e-05,
|
| 2347 |
+
"loss": 454.87333984375,
|
| 2348 |
+
"step": 6660
|
| 2349 |
+
},
|
| 2350 |
+
{
|
| 2351 |
+
"epoch": 0.4057006098035457,
|
| 2352 |
+
"grad_norm": 116.46212768554688,
|
| 2353 |
+
"learning_rate": 4.999932554490078e-05,
|
| 2354 |
+
"loss": 454.346142578125,
|
| 2355 |
+
"step": 6680
|
| 2356 |
+
},
|
| 2357 |
+
{
|
| 2358 |
+
"epoch": 0.40691528228798746,
|
| 2359 |
+
"grad_norm": 136.41693115234375,
|
| 2360 |
+
"learning_rate": 4.999920370099529e-05,
|
| 2361 |
+
"loss": 451.74169921875,
|
| 2362 |
+
"step": 6700
|
| 2363 |
+
},
|
| 2364 |
+
{
|
| 2365 |
+
"epoch": 0.4081299547724292,
|
| 2366 |
+
"grad_norm": 111.75225830078125,
|
| 2367 |
+
"learning_rate": 4.999907174578565e-05,
|
| 2368 |
+
"loss": 451.062451171875,
|
| 2369 |
+
"step": 6720
|
| 2370 |
+
},
|
| 2371 |
+
{
|
| 2372 |
+
"epoch": 0.409344627256871,
|
| 2373 |
+
"grad_norm": 117.8348617553711,
|
| 2374 |
+
"learning_rate": 4.999892967932522e-05,
|
| 2375 |
+
"loss": 448.330615234375,
|
| 2376 |
+
"step": 6740
|
| 2377 |
+
},
|
| 2378 |
+
{
|
| 2379 |
+
"epoch": 0.4105592997413127,
|
| 2380 |
+
"grad_norm": 99.80577087402344,
|
| 2381 |
+
"learning_rate": 4.999877750167147e-05,
|
| 2382 |
+
"loss": 450.1162109375,
|
| 2383 |
+
"step": 6760
|
| 2384 |
+
},
|
| 2385 |
+
{
|
| 2386 |
+
"epoch": 0.41177397222575446,
|
| 2387 |
+
"grad_norm": 142.496826171875,
|
| 2388 |
+
"learning_rate": 4.999861521288595e-05,
|
| 2389 |
+
"loss": 451.98525390625,
|
| 2390 |
+
"step": 6780
|
| 2391 |
+
},
|
| 2392 |
+
{
|
| 2393 |
+
"epoch": 0.41298864471019625,
|
| 2394 |
+
"grad_norm": 112.31470489501953,
|
| 2395 |
+
"learning_rate": 4.999844281303429e-05,
|
| 2396 |
+
"loss": 454.02314453125,
|
| 2397 |
+
"step": 6800
|
| 2398 |
+
},
|
| 2399 |
+
{
|
| 2400 |
+
"epoch": 0.414203317194638,
|
| 2401 |
+
"grad_norm": 130.9207000732422,
|
| 2402 |
+
"learning_rate": 4.999999919106988e-05,
|
| 2403 |
+
"loss": 434.83544921875,
|
| 2404 |
+
"step": 6820
|
| 2405 |
+
},
|
| 2406 |
+
{
|
| 2407 |
+
"epoch": 0.4154179896790797,
|
| 2408 |
+
"grad_norm": 107.9826889038086,
|
| 2409 |
+
"learning_rate": 4.999999009060662e-05,
|
| 2410 |
+
"loss": 431.314208984375,
|
| 2411 |
+
"step": 6840
|
| 2412 |
+
},
|
| 2413 |
+
{
|
| 2414 |
+
"epoch": 0.41663266216352146,
|
| 2415 |
+
"grad_norm": 112.3042221069336,
|
| 2416 |
+
"learning_rate": 4.9999970878521126e-05,
|
| 2417 |
+
"loss": 445.6931640625,
|
| 2418 |
+
"step": 6860
|
| 2419 |
+
},
|
| 2420 |
+
{
|
| 2421 |
+
"epoch": 0.41784733464796325,
|
| 2422 |
+
"grad_norm": 114.62458038330078,
|
| 2423 |
+
"learning_rate": 4.99999415548212e-05,
|
| 2424 |
+
"loss": 438.76865234375,
|
| 2425 |
+
"step": 6880
|
| 2426 |
+
},
|
| 2427 |
+
{
|
| 2428 |
+
"epoch": 0.419062007132405,
|
| 2429 |
+
"grad_norm": 105.3339614868164,
|
| 2430 |
+
"learning_rate": 4.999990211951866e-05,
|
| 2431 |
+
"loss": 433.6736328125,
|
| 2432 |
+
"step": 6900
|
| 2433 |
+
},
|
| 2434 |
+
{
|
| 2435 |
+
"epoch": 0.4202766796168467,
|
| 2436 |
+
"grad_norm": 108.74405670166016,
|
| 2437 |
+
"learning_rate": 4.999985257262949e-05,
|
| 2438 |
+
"loss": 429.842138671875,
|
| 2439 |
+
"step": 6920
|
| 2440 |
+
},
|
| 2441 |
+
{
|
| 2442 |
+
"epoch": 0.4214913521012885,
|
| 2443 |
+
"grad_norm": 122.67109680175781,
|
| 2444 |
+
"learning_rate": 4.999979291417372e-05,
|
| 2445 |
+
"loss": 436.727734375,
|
| 2446 |
+
"step": 6940
|
| 2447 |
+
},
|
| 2448 |
+
{
|
| 2449 |
+
"epoch": 0.42270602458573026,
|
| 2450 |
+
"grad_norm": 112.16912078857422,
|
| 2451 |
+
"learning_rate": 4.9999723144175475e-05,
|
| 2452 |
+
"loss": 431.93212890625,
|
| 2453 |
+
"step": 6960
|
| 2454 |
+
},
|
| 2455 |
+
{
|
| 2456 |
+
"epoch": 0.423920697070172,
|
| 2457 |
+
"grad_norm": 108.9756088256836,
|
| 2458 |
+
"learning_rate": 4.999964326266298e-05,
|
| 2459 |
+
"loss": 423.93974609375,
|
| 2460 |
+
"step": 6980
|
| 2461 |
+
},
|
| 2462 |
+
{
|
| 2463 |
+
"epoch": 0.4251353695546138,
|
| 2464 |
+
"grad_norm": 109.26213073730469,
|
| 2465 |
+
"learning_rate": 4.9999553269668544e-05,
|
| 2466 |
+
"loss": 429.35048828125,
|
| 2467 |
+
"step": 7000
|
| 2468 |
+
},
|
| 2469 |
+
{
|
| 2470 |
+
"epoch": 0.4263500420390555,
|
| 2471 |
+
"grad_norm": 128.8697052001953,
|
| 2472 |
+
"learning_rate": 4.999945316522856e-05,
|
| 2473 |
+
"loss": 432.916650390625,
|
| 2474 |
+
"step": 7020
|
| 2475 |
+
},
|
| 2476 |
+
{
|
| 2477 |
+
"epoch": 0.42756471452349726,
|
| 2478 |
+
"grad_norm": 116.16033172607422,
|
| 2479 |
+
"learning_rate": 4.999934294938352e-05,
|
| 2480 |
+
"loss": 425.26083984375,
|
| 2481 |
+
"step": 7040
|
| 2482 |
+
},
|
| 2483 |
+
{
|
| 2484 |
+
"epoch": 0.42877938700793905,
|
| 2485 |
+
"grad_norm": 112.03089141845703,
|
| 2486 |
+
"learning_rate": 4.9999222622178e-05,
|
| 2487 |
+
"loss": 426.267236328125,
|
| 2488 |
+
"step": 7060
|
| 2489 |
+
},
|
| 2490 |
+
{
|
| 2491 |
+
"epoch": 0.4299940594923808,
|
| 2492 |
+
"grad_norm": 113.0052261352539,
|
| 2493 |
+
"learning_rate": 4.9999092183660675e-05,
|
| 2494 |
+
"loss": 425.38056640625,
|
| 2495 |
+
"step": 7080
|
| 2496 |
+
},
|
| 2497 |
+
{
|
| 2498 |
+
"epoch": 0.4312087319768225,
|
| 2499 |
+
"grad_norm": 103.9638900756836,
|
| 2500 |
+
"learning_rate": 4.999895163388431e-05,
|
| 2501 |
+
"loss": 421.51259765625,
|
| 2502 |
+
"step": 7100
|
| 2503 |
+
},
|
| 2504 |
+
{
|
| 2505 |
+
"epoch": 0.43242340446126426,
|
| 2506 |
+
"grad_norm": 105.279541015625,
|
| 2507 |
+
"learning_rate": 4.9998800972905736e-05,
|
| 2508 |
+
"loss": 426.773291015625,
|
| 2509 |
+
"step": 7120
|
| 2510 |
+
},
|
| 2511 |
+
{
|
| 2512 |
+
"epoch": 0.43363807694570605,
|
| 2513 |
+
"grad_norm": 111.30103302001953,
|
| 2514 |
+
"learning_rate": 4.999864020078589e-05,
|
| 2515 |
+
"loss": 426.06201171875,
|
| 2516 |
+
"step": 7140
|
| 2517 |
+
},
|
| 2518 |
+
{
|
| 2519 |
+
"epoch": 0.4348527494301478,
|
| 2520 |
+
"grad_norm": 99.83489227294922,
|
| 2521 |
+
"learning_rate": 4.999846931758981e-05,
|
| 2522 |
+
"loss": 421.738671875,
|
| 2523 |
+
"step": 7160
|
| 2524 |
+
},
|
| 2525 |
+
{
|
| 2526 |
+
"epoch": 0.4360674219145895,
|
| 2527 |
+
"grad_norm": 111.20061492919922,
|
| 2528 |
+
"learning_rate": 4.999828832338661e-05,
|
| 2529 |
+
"loss": 426.967919921875,
|
| 2530 |
+
"step": 7180
|
| 2531 |
+
},
|
| 2532 |
+
{
|
| 2533 |
+
"epoch": 0.4372820943990313,
|
| 2534 |
+
"grad_norm": 116.54100036621094,
|
| 2535 |
+
"learning_rate": 4.999809721824949e-05,
|
| 2536 |
+
"loss": 419.972900390625,
|
| 2537 |
+
"step": 7200
|
| 2538 |
+
},
|
| 2539 |
+
{
|
| 2540 |
+
"epoch": 0.43849676688347305,
|
| 2541 |
+
"grad_norm": 117.1341781616211,
|
| 2542 |
+
"learning_rate": 4.999789600225574e-05,
|
| 2543 |
+
"loss": 412.59365234375,
|
| 2544 |
+
"step": 7220
|
| 2545 |
+
},
|
| 2546 |
+
{
|
| 2547 |
+
"epoch": 0.4397114393679148,
|
| 2548 |
+
"grad_norm": 104.19764709472656,
|
| 2549 |
+
"learning_rate": 4.999768467548675e-05,
|
| 2550 |
+
"loss": 419.56875,
|
| 2551 |
+
"step": 7240
|
| 2552 |
+
},
|
| 2553 |
+
{
|
| 2554 |
+
"epoch": 0.4409261118523566,
|
| 2555 |
+
"grad_norm": 104.99066925048828,
|
| 2556 |
+
"learning_rate": 4.999746323802802e-05,
|
| 2557 |
+
"loss": 420.034375,
|
| 2558 |
+
"step": 7260
|
| 2559 |
+
},
|
| 2560 |
+
{
|
| 2561 |
+
"epoch": 0.4421407843367983,
|
| 2562 |
+
"grad_norm": 117.05961608886719,
|
| 2563 |
+
"learning_rate": 4.999723168996907e-05,
|
| 2564 |
+
"loss": 421.9271484375,
|
| 2565 |
+
"step": 7280
|
| 2566 |
+
},
|
| 2567 |
+
{
|
| 2568 |
+
"epoch": 0.44335545682124006,
|
| 2569 |
+
"grad_norm": 140.603271484375,
|
| 2570 |
+
"learning_rate": 4.9996990031403586e-05,
|
| 2571 |
+
"loss": 414.878369140625,
|
| 2572 |
+
"step": 7300
|
| 2573 |
+
},
|
| 2574 |
+
{
|
| 2575 |
+
"epoch": 0.4445701293056818,
|
| 2576 |
+
"grad_norm": 114.2315902709961,
|
| 2577 |
+
"learning_rate": 4.999673826242929e-05,
|
| 2578 |
+
"loss": 415.910205078125,
|
| 2579 |
+
"step": 7320
|
| 2580 |
+
},
|
| 2581 |
+
{
|
| 2582 |
+
"epoch": 0.4457848017901236,
|
| 2583 |
+
"grad_norm": 107.20604705810547,
|
| 2584 |
+
"learning_rate": 4.9996476383148024e-05,
|
| 2585 |
+
"loss": 417.096337890625,
|
| 2586 |
+
"step": 7340
|
| 2587 |
+
},
|
| 2588 |
+
{
|
| 2589 |
+
"epoch": 0.4469994742745653,
|
| 2590 |
+
"grad_norm": 120.31047821044922,
|
| 2591 |
+
"learning_rate": 4.9996204393665716e-05,
|
| 2592 |
+
"loss": 422.905029296875,
|
| 2593 |
+
"step": 7360
|
| 2594 |
+
},
|
| 2595 |
+
{
|
| 2596 |
+
"epoch": 0.44821414675900706,
|
| 2597 |
+
"grad_norm": 118.30343627929688,
|
| 2598 |
+
"learning_rate": 4.999592229409236e-05,
|
| 2599 |
+
"loss": 418.848193359375,
|
| 2600 |
+
"step": 7380
|
| 2601 |
+
},
|
| 2602 |
+
{
|
| 2603 |
+
"epoch": 0.44942881924344885,
|
| 2604 |
+
"grad_norm": 106.66614532470703,
|
| 2605 |
+
"learning_rate": 4.999563008454207e-05,
|
| 2606 |
+
"loss": 419.152685546875,
|
| 2607 |
+
"step": 7400
|
| 2608 |
+
},
|
| 2609 |
+
{
|
| 2610 |
+
"epoch": 0.4506434917278906,
|
| 2611 |
+
"grad_norm": 109.3532943725586,
|
| 2612 |
+
"learning_rate": 4.999532776513301e-05,
|
| 2613 |
+
"loss": 415.478466796875,
|
| 2614 |
+
"step": 7420
|
| 2615 |
+
},
|
| 2616 |
+
{
|
| 2617 |
+
"epoch": 0.4518581642123323,
|
| 2618 |
+
"grad_norm": 122.94757843017578,
|
| 2619 |
+
"learning_rate": 4.9995015335987497e-05,
|
| 2620 |
+
"loss": 415.234130859375,
|
| 2621 |
+
"step": 7440
|
| 2622 |
+
},
|
| 2623 |
+
{
|
| 2624 |
+
"epoch": 0.4530728366967741,
|
| 2625 |
+
"grad_norm": 108.08473205566406,
|
| 2626 |
+
"learning_rate": 4.999469279723187e-05,
|
| 2627 |
+
"loss": 411.04873046875,
|
| 2628 |
+
"step": 7460
|
| 2629 |
+
},
|
| 2630 |
+
{
|
| 2631 |
+
"epoch": 0.45428750918121585,
|
| 2632 |
+
"grad_norm": 104.21884155273438,
|
| 2633 |
+
"learning_rate": 4.9994360148996586e-05,
|
| 2634 |
+
"loss": 414.63203125,
|
| 2635 |
+
"step": 7480
|
| 2636 |
+
},
|
| 2637 |
+
{
|
| 2638 |
+
"epoch": 0.4555021816656576,
|
| 2639 |
+
"grad_norm": 108.63338470458984,
|
| 2640 |
+
"learning_rate": 4.999401739141619e-05,
|
| 2641 |
+
"loss": 418.697509765625,
|
| 2642 |
+
"step": 7500
|
| 2643 |
+
},
|
| 2644 |
+
{
|
| 2645 |
+
"epoch": 0.4567168541500994,
|
| 2646 |
+
"grad_norm": 119.93159484863281,
|
| 2647 |
+
"learning_rate": 4.999366452462933e-05,
|
| 2648 |
+
"loss": 412.669287109375,
|
| 2649 |
+
"step": 7520
|
| 2650 |
+
},
|
| 2651 |
+
{
|
| 2652 |
+
"epoch": 0.4579315266345411,
|
| 2653 |
+
"grad_norm": 113.58409118652344,
|
| 2654 |
+
"learning_rate": 4.9993301548778705e-05,
|
| 2655 |
+
"loss": 412.94091796875,
|
| 2656 |
+
"step": 7540
|
| 2657 |
+
},
|
| 2658 |
+
{
|
| 2659 |
+
"epoch": 0.45914619911898286,
|
| 2660 |
+
"grad_norm": 129.66651916503906,
|
| 2661 |
+
"learning_rate": 4.9992928464011144e-05,
|
| 2662 |
+
"loss": 411.043701171875,
|
| 2663 |
+
"step": 7560
|
| 2664 |
+
},
|
| 2665 |
+
{
|
| 2666 |
+
"epoch": 0.4603608716034246,
|
| 2667 |
+
"grad_norm": 98.4170913696289,
|
| 2668 |
+
"learning_rate": 4.9992545270477545e-05,
|
| 2669 |
+
"loss": 401.46005859375,
|
| 2670 |
+
"step": 7580
|
| 2671 |
+
},
|
| 2672 |
+
{
|
| 2673 |
+
"epoch": 0.4615755440878664,
|
| 2674 |
+
"grad_norm": 174.97955322265625,
|
| 2675 |
+
"learning_rate": 4.999215196833289e-05,
|
| 2676 |
+
"loss": 419.571337890625,
|
| 2677 |
+
"step": 7600
|
| 2678 |
+
},
|
| 2679 |
+
{
|
| 2680 |
+
"epoch": 0.4627902165723081,
|
| 2681 |
+
"grad_norm": 116.11487579345703,
|
| 2682 |
+
"learning_rate": 4.999174855773626e-05,
|
| 2683 |
+
"loss": 409.20986328125,
|
| 2684 |
+
"step": 7620
|
| 2685 |
+
},
|
| 2686 |
+
{
|
| 2687 |
+
"epoch": 0.46400488905674986,
|
| 2688 |
+
"grad_norm": 123.58662414550781,
|
| 2689 |
+
"learning_rate": 4.9991335038850826e-05,
|
| 2690 |
+
"loss": 418.3943359375,
|
| 2691 |
+
"step": 7640
|
| 2692 |
+
},
|
| 2693 |
+
{
|
| 2694 |
+
"epoch": 0.46521956154119165,
|
| 2695 |
+
"grad_norm": 118.61854553222656,
|
| 2696 |
+
"learning_rate": 4.9990911411843824e-05,
|
| 2697 |
+
"loss": 410.848095703125,
|
| 2698 |
+
"step": 7660
|
| 2699 |
+
},
|
| 2700 |
+
{
|
| 2701 |
+
"epoch": 0.4664342340256334,
|
| 2702 |
+
"grad_norm": 145.44140625,
|
| 2703 |
+
"learning_rate": 4.999047767688661e-05,
|
| 2704 |
+
"loss": 415.935986328125,
|
| 2705 |
+
"step": 7680
|
| 2706 |
+
},
|
| 2707 |
+
{
|
| 2708 |
+
"epoch": 0.4676489065100751,
|
| 2709 |
+
"grad_norm": 137.12423706054688,
|
| 2710 |
+
"learning_rate": 4.999003383415462e-05,
|
| 2711 |
+
"loss": 409.5922607421875,
|
| 2712 |
+
"step": 7700
|
| 2713 |
+
},
|
| 2714 |
+
{
|
| 2715 |
+
"epoch": 0.2930397136979222,
|
| 2716 |
+
"grad_norm": 100.56342315673828,
|
| 2717 |
+
"learning_rate": 4.999999676427957e-05,
|
| 2718 |
+
"loss": 408.933984375,
|
| 2719 |
+
"step": 7720
|
| 2720 |
+
},
|
| 2721 |
+
{
|
| 2722 |
+
"epoch": 0.29379888394066295,
|
| 2723 |
+
"grad_norm": 134.54612731933594,
|
| 2724 |
+
"learning_rate": 4.999998361916675e-05,
|
| 2725 |
+
"loss": 396.3473876953125,
|
| 2726 |
+
"step": 7740
|
| 2727 |
+
},
|
| 2728 |
+
{
|
| 2729 |
+
"epoch": 0.2945580541834037,
|
| 2730 |
+
"grad_norm": 117.5848159790039,
|
| 2731 |
+
"learning_rate": 4.999996036243432e-05,
|
| 2732 |
+
"loss": 399.0041259765625,
|
| 2733 |
+
"step": 7760
|
| 2734 |
+
},
|
| 2735 |
+
{
|
| 2736 |
+
"epoch": 0.2953172244261444,
|
| 2737 |
+
"grad_norm": 102.51106262207031,
|
| 2738 |
+
"learning_rate": 4.999992699409169e-05,
|
| 2739 |
+
"loss": 412.077001953125,
|
| 2740 |
+
"step": 7780
|
| 2741 |
+
},
|
| 2742 |
+
{
|
| 2743 |
+
"epoch": 0.29607639466888513,
|
| 2744 |
+
"grad_norm": 113.85884094238281,
|
| 2745 |
+
"learning_rate": 4.9999883514152356e-05,
|
| 2746 |
+
"loss": 409.868994140625,
|
| 2747 |
+
"step": 7800
|
| 2748 |
+
},
|
| 2749 |
+
{
|
| 2750 |
+
"epoch": 0.2968355649116259,
|
| 2751 |
+
"grad_norm": 104.08061218261719,
|
| 2752 |
+
"learning_rate": 4.999982992263391e-05,
|
| 2753 |
+
"loss": 409.793896484375,
|
| 2754 |
+
"step": 7820
|
| 2755 |
+
},
|
| 2756 |
+
{
|
| 2757 |
+
"epoch": 0.2975947351543666,
|
| 2758 |
+
"grad_norm": 129.54656982421875,
|
| 2759 |
+
"learning_rate": 4.999976621955803e-05,
|
| 2760 |
+
"loss": 406.743408203125,
|
| 2761 |
+
"step": 7840
|
| 2762 |
+
},
|
| 2763 |
+
{
|
| 2764 |
+
"epoch": 0.2983539053971073,
|
| 2765 |
+
"grad_norm": 111.32732391357422,
|
| 2766 |
+
"learning_rate": 4.9999692404950473e-05,
|
| 2767 |
+
"loss": 405.322509765625,
|
| 2768 |
+
"step": 7860
|
| 2769 |
+
},
|
| 2770 |
+
{
|
| 2771 |
+
"epoch": 0.29911307563984807,
|
| 2772 |
+
"grad_norm": 121.55047607421875,
|
| 2773 |
+
"learning_rate": 4.9999608478841085e-05,
|
| 2774 |
+
"loss": 409.610107421875,
|
| 2775 |
+
"step": 7880
|
| 2776 |
+
},
|
| 2777 |
+
{
|
| 2778 |
+
"epoch": 0.29987224588258876,
|
| 2779 |
+
"grad_norm": 127.51071166992188,
|
| 2780 |
+
"learning_rate": 4.999951444126384e-05,
|
| 2781 |
+
"loss": 408.4365966796875,
|
| 2782 |
+
"step": 7900
|
| 2783 |
+
},
|
| 2784 |
+
{
|
| 2785 |
+
"epoch": 0.3006314161253295,
|
| 2786 |
+
"grad_norm": 105.90245819091797,
|
| 2787 |
+
"learning_rate": 4.999941029225674e-05,
|
| 2788 |
+
"loss": 400.70888671875,
|
| 2789 |
+
"step": 7920
|
| 2790 |
+
},
|
| 2791 |
+
{
|
| 2792 |
+
"epoch": 0.30139058636807026,
|
| 2793 |
+
"grad_norm": 115.70623779296875,
|
| 2794 |
+
"learning_rate": 4.9999296031861934e-05,
|
| 2795 |
+
"loss": 403.0365478515625,
|
| 2796 |
+
"step": 7940
|
| 2797 |
+
},
|
| 2798 |
+
{
|
| 2799 |
+
"epoch": 0.302149756610811,
|
| 2800 |
+
"grad_norm": 108.50679779052734,
|
| 2801 |
+
"learning_rate": 4.9999171660125635e-05,
|
| 2802 |
+
"loss": 397.7148681640625,
|
| 2803 |
+
"step": 7960
|
| 2804 |
+
}
|
| 2805 |
+
],
|
| 2806 |
+
"logging_steps": 20,
|
| 2807 |
+
"max_steps": 98796,
|
| 2808 |
+
"num_input_tokens_seen": 0,
|
| 2809 |
+
"num_train_epochs": 4,
|
| 2810 |
+
"save_steps": 1000000000,
|
| 2811 |
+
"stateful_callbacks": {
|
| 2812 |
+
"TrainerControl": {
|
| 2813 |
+
"args": {
|
| 2814 |
+
"should_epoch_stop": false,
|
| 2815 |
+
"should_evaluate": false,
|
| 2816 |
+
"should_log": false,
|
| 2817 |
+
"should_save": true,
|
| 2818 |
+
"should_training_stop": false
|
| 2819 |
+
},
|
| 2820 |
+
"attributes": {}
|
| 2821 |
+
}
|
| 2822 |
+
},
|
| 2823 |
+
"total_flos": 2.1460855618215936e+17,
|
| 2824 |
+
"train_batch_size": 1,
|
| 2825 |
+
"trial_name": null,
|
| 2826 |
+
"trial_params": null
|
| 2827 |
+
}
|