Training in progress, step 18000, checkpoint
Browse files- checkpoint-18000/config.json +47 -0
- checkpoint-18000/generation_config.json +10 -0
- checkpoint-18000/model.safetensors +3 -0
- checkpoint-18000/optimizer.pt +3 -0
- checkpoint-18000/rng_state.pth +3 -0
- checkpoint-18000/scheduler.pt +3 -0
- checkpoint-18000/trainer_state.json +2602 -0
- checkpoint-18000/training_args.bin +3 -0
checkpoint-18000/config.json
ADDED
|
@@ -0,0 +1,47 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"Qwen3ForCausalLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_bias": false,
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"bos_token_id": 0,
|
| 8 |
+
"dtype": "float32",
|
| 9 |
+
"eos_token_id": 2,
|
| 10 |
+
"head_dim": 64,
|
| 11 |
+
"hidden_act": "silu",
|
| 12 |
+
"hidden_size": 768,
|
| 13 |
+
"initializer_range": 0.02,
|
| 14 |
+
"intermediate_size": 2048,
|
| 15 |
+
"layer_types": [
|
| 16 |
+
"full_attention",
|
| 17 |
+
"full_attention",
|
| 18 |
+
"full_attention",
|
| 19 |
+
"full_attention",
|
| 20 |
+
"full_attention",
|
| 21 |
+
"full_attention",
|
| 22 |
+
"full_attention",
|
| 23 |
+
"full_attention",
|
| 24 |
+
"full_attention",
|
| 25 |
+
"full_attention",
|
| 26 |
+
"full_attention",
|
| 27 |
+
"full_attention"
|
| 28 |
+
],
|
| 29 |
+
"max_position_embeddings": 2048,
|
| 30 |
+
"max_window_layers": 28,
|
| 31 |
+
"model_type": "qwen3",
|
| 32 |
+
"num_attention_heads": 12,
|
| 33 |
+
"num_hidden_layers": 12,
|
| 34 |
+
"num_key_value_heads": 4,
|
| 35 |
+
"pad_token_id": 1,
|
| 36 |
+
"rms_norm_eps": 1e-06,
|
| 37 |
+
"rope_parameters": {
|
| 38 |
+
"rope_theta": 10000,
|
| 39 |
+
"rope_type": "default"
|
| 40 |
+
},
|
| 41 |
+
"sliding_window": null,
|
| 42 |
+
"tie_word_embeddings": true,
|
| 43 |
+
"transformers_version": "5.14.1",
|
| 44 |
+
"use_cache": false,
|
| 45 |
+
"use_sliding_window": false,
|
| 46 |
+
"vocab_size": 32768
|
| 47 |
+
}
|
checkpoint-18000/generation_config.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 0,
|
| 4 |
+
"eos_token_id": 2,
|
| 5 |
+
"output_attentions": false,
|
| 6 |
+
"output_hidden_states": false,
|
| 7 |
+
"pad_token_id": 1,
|
| 8 |
+
"transformers_version": "5.14.1",
|
| 9 |
+
"use_cache": true
|
| 10 |
+
}
|
checkpoint-18000/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bfd681a39364e6440a0397c35fe1cd1cafd2b1e6a6fb116d5d4d1e7d60fa761a
|
| 3 |
+
size 402750928
|
checkpoint-18000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ab385357da73c3d26459b9a4ceeec7cc5da99697ff88ad036c857efc5780ab5a
|
| 3 |
+
size 805587275
|
checkpoint-18000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7ffffd41cb5466a23271a8c6303a885c8e75cf7a03496c0a7cb7f448827155d8
|
| 3 |
+
size 14645
|
checkpoint-18000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:033a150db651b20df4d58c899bfbbf307e0775cebeee0e9f35999aedf205f306
|
| 3 |
+
size 1465
|
checkpoint-18000/trainer_state.json
ADDED
|
@@ -0,0 +1,2602 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.157286373156459,
|
| 6 |
+
"eval_steps": 3000,
|
| 7 |
+
"global_step": 18000,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.00043690659210127495,
|
| 14 |
+
"grad_norm": 1.8826903104782104,
|
| 15 |
+
"learning_rate": 2.4500000000000003e-05,
|
| 16 |
+
"loss": 9.762939453125,
|
| 17 |
+
"step": 50
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.0008738131842025499,
|
| 21 |
+
"grad_norm": 0.9995923042297363,
|
| 22 |
+
"learning_rate": 4.9500000000000004e-05,
|
| 23 |
+
"loss": 8.305597534179688,
|
| 24 |
+
"step": 100
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.001310719776303825,
|
| 28 |
+
"grad_norm": 0.9061496257781982,
|
| 29 |
+
"learning_rate": 7.45e-05,
|
| 30 |
+
"loss": 7.173495483398438,
|
| 31 |
+
"step": 150
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.0017476263684050998,
|
| 35 |
+
"grad_norm": 1.6024847030639648,
|
| 36 |
+
"learning_rate": 9.95e-05,
|
| 37 |
+
"loss": 6.624603271484375,
|
| 38 |
+
"step": 200
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.002184532960506375,
|
| 42 |
+
"grad_norm": 1.2123334407806396,
|
| 43 |
+
"learning_rate": 0.0001245,
|
| 44 |
+
"loss": 6.243399658203125,
|
| 45 |
+
"step": 250
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.00262143955260765,
|
| 49 |
+
"grad_norm": 1.0432322025299072,
|
| 50 |
+
"learning_rate": 0.0001495,
|
| 51 |
+
"loss": 5.9879296875,
|
| 52 |
+
"step": 300
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.0030583461447089245,
|
| 56 |
+
"grad_norm": 0.9280146360397339,
|
| 57 |
+
"learning_rate": 0.00017449999999999999,
|
| 58 |
+
"loss": 5.776607666015625,
|
| 59 |
+
"step": 350
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.0034952527368101996,
|
| 63 |
+
"grad_norm": 1.0289900302886963,
|
| 64 |
+
"learning_rate": 0.00019950000000000002,
|
| 65 |
+
"loss": 5.5920703125,
|
| 66 |
+
"step": 400
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.003932159328911475,
|
| 70 |
+
"grad_norm": 1.0593407154083252,
|
| 71 |
+
"learning_rate": 0.0002245,
|
| 72 |
+
"loss": 5.437271728515625,
|
| 73 |
+
"step": 450
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.00436906592101275,
|
| 77 |
+
"grad_norm": 1.0122662782669067,
|
| 78 |
+
"learning_rate": 0.0002495,
|
| 79 |
+
"loss": 5.287777709960937,
|
| 80 |
+
"step": 500
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.004805972513114025,
|
| 84 |
+
"grad_norm": 0.7662743926048279,
|
| 85 |
+
"learning_rate": 0.0002745,
|
| 86 |
+
"loss": 5.155453491210937,
|
| 87 |
+
"step": 550
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.0052428791052153,
|
| 91 |
+
"grad_norm": 0.7941539287567139,
|
| 92 |
+
"learning_rate": 0.0002995,
|
| 93 |
+
"loss": 5.030631713867187,
|
| 94 |
+
"step": 600
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.005679785697316574,
|
| 98 |
+
"grad_norm": 0.6195730566978455,
|
| 99 |
+
"learning_rate": 0.00032450000000000003,
|
| 100 |
+
"loss": 4.909389953613282,
|
| 101 |
+
"step": 650
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.006116692289417849,
|
| 105 |
+
"grad_norm": 0.6587674617767334,
|
| 106 |
+
"learning_rate": 0.0003495,
|
| 107 |
+
"loss": 4.786273498535156,
|
| 108 |
+
"step": 700
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.006553598881519124,
|
| 112 |
+
"grad_norm": 0.7233771681785583,
|
| 113 |
+
"learning_rate": 0.0003745,
|
| 114 |
+
"loss": 4.6648342895507815,
|
| 115 |
+
"step": 750
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.006990505473620399,
|
| 119 |
+
"grad_norm": 0.5489601492881775,
|
| 120 |
+
"learning_rate": 0.0003995,
|
| 121 |
+
"loss": 4.555048522949218,
|
| 122 |
+
"step": 800
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.007427412065721674,
|
| 126 |
+
"grad_norm": 0.6154086589813232,
|
| 127 |
+
"learning_rate": 0.0004245,
|
| 128 |
+
"loss": 4.461051330566407,
|
| 129 |
+
"step": 850
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.00786431865782295,
|
| 133 |
+
"grad_norm": 0.41333407163619995,
|
| 134 |
+
"learning_rate": 0.00044950000000000003,
|
| 135 |
+
"loss": 4.374936828613281,
|
| 136 |
+
"step": 900
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.008301225249924224,
|
| 140 |
+
"grad_norm": 0.5036063194274902,
|
| 141 |
+
"learning_rate": 0.0004745,
|
| 142 |
+
"loss": 4.3033209228515625,
|
| 143 |
+
"step": 950
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.0087381318420255,
|
| 147 |
+
"grad_norm": 0.3687891364097595,
|
| 148 |
+
"learning_rate": 0.0004995,
|
| 149 |
+
"loss": 4.233715209960938,
|
| 150 |
+
"step": 1000
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.009175038434126774,
|
| 154 |
+
"grad_norm": 0.5433846116065979,
|
| 155 |
+
"learning_rate": 0.0005245,
|
| 156 |
+
"loss": 4.186253967285157,
|
| 157 |
+
"step": 1050
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.00961194502622805,
|
| 161 |
+
"grad_norm": 0.439351350069046,
|
| 162 |
+
"learning_rate": 0.0005495,
|
| 163 |
+
"loss": 4.139383850097656,
|
| 164 |
+
"step": 1100
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.010048851618329325,
|
| 168 |
+
"grad_norm": 0.36215880513191223,
|
| 169 |
+
"learning_rate": 0.0005745,
|
| 170 |
+
"loss": 4.08899658203125,
|
| 171 |
+
"step": 1150
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.0104857582104306,
|
| 175 |
+
"grad_norm": 0.3970701992511749,
|
| 176 |
+
"learning_rate": 0.0005995000000000001,
|
| 177 |
+
"loss": 4.061238403320313,
|
| 178 |
+
"step": 1200
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.010922664802531873,
|
| 182 |
+
"grad_norm": 0.3410918712615967,
|
| 183 |
+
"learning_rate": 0.0006245000000000001,
|
| 184 |
+
"loss": 4.026336669921875,
|
| 185 |
+
"step": 1250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.011359571394633148,
|
| 189 |
+
"grad_norm": 0.3047930598258972,
|
| 190 |
+
"learning_rate": 0.0006495,
|
| 191 |
+
"loss": 3.9928851318359375,
|
| 192 |
+
"step": 1300
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.011796477986734423,
|
| 196 |
+
"grad_norm": 0.2819700241088867,
|
| 197 |
+
"learning_rate": 0.0006745,
|
| 198 |
+
"loss": 3.9584146118164063,
|
| 199 |
+
"step": 1350
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.012233384578835698,
|
| 203 |
+
"grad_norm": 0.29095014929771423,
|
| 204 |
+
"learning_rate": 0.0006995,
|
| 205 |
+
"loss": 3.9291363525390626,
|
| 206 |
+
"step": 1400
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.012670291170936973,
|
| 210 |
+
"grad_norm": 0.2987957298755646,
|
| 211 |
+
"learning_rate": 0.0007245000000000001,
|
| 212 |
+
"loss": 3.9090225219726564,
|
| 213 |
+
"step": 1450
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.013107197763038248,
|
| 217 |
+
"grad_norm": 0.28419050574302673,
|
| 218 |
+
"learning_rate": 0.0007495000000000001,
|
| 219 |
+
"loss": 3.890744934082031,
|
| 220 |
+
"step": 1500
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.013544104355139523,
|
| 224 |
+
"grad_norm": 0.3039053976535797,
|
| 225 |
+
"learning_rate": 0.0007745,
|
| 226 |
+
"loss": 3.861073303222656,
|
| 227 |
+
"step": 1550
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.013981010947240798,
|
| 231 |
+
"grad_norm": 0.25676754117012024,
|
| 232 |
+
"learning_rate": 0.0007995,
|
| 233 |
+
"loss": 3.8428253173828124,
|
| 234 |
+
"step": 1600
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.014417917539342073,
|
| 238 |
+
"grad_norm": 0.2573896646499634,
|
| 239 |
+
"learning_rate": 0.0008245,
|
| 240 |
+
"loss": 3.8260653686523436,
|
| 241 |
+
"step": 1650
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.014854824131443348,
|
| 245 |
+
"grad_norm": 0.272127240896225,
|
| 246 |
+
"learning_rate": 0.0008495000000000001,
|
| 247 |
+
"loss": 3.8006341552734373,
|
| 248 |
+
"step": 1700
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.015291730723544623,
|
| 252 |
+
"grad_norm": 0.23205944895744324,
|
| 253 |
+
"learning_rate": 0.0008745000000000001,
|
| 254 |
+
"loss": 3.7862649536132813,
|
| 255 |
+
"step": 1750
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.0157286373156459,
|
| 259 |
+
"grad_norm": 0.22847425937652588,
|
| 260 |
+
"learning_rate": 0.0008995,
|
| 261 |
+
"loss": 3.771934814453125,
|
| 262 |
+
"step": 1800
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.016165543907747174,
|
| 266 |
+
"grad_norm": 0.2651786208152771,
|
| 267 |
+
"learning_rate": 0.0009245,
|
| 268 |
+
"loss": 3.76009521484375,
|
| 269 |
+
"step": 1850
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.01660245049984845,
|
| 273 |
+
"grad_norm": 0.24903950095176697,
|
| 274 |
+
"learning_rate": 0.0009495,
|
| 275 |
+
"loss": 3.7454238891601563,
|
| 276 |
+
"step": 1900
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.017039357091949724,
|
| 280 |
+
"grad_norm": 0.24524764716625214,
|
| 281 |
+
"learning_rate": 0.0009745000000000001,
|
| 282 |
+
"loss": 3.735928955078125,
|
| 283 |
+
"step": 1950
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.017476263684051,
|
| 287 |
+
"grad_norm": 0.22325558960437775,
|
| 288 |
+
"learning_rate": 0.0009995000000000002,
|
| 289 |
+
"loss": 3.7209417724609377,
|
| 290 |
+
"step": 2000
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.017913170276152274,
|
| 294 |
+
"grad_norm": 0.23860357701778412,
|
| 295 |
+
"learning_rate": 0.001,
|
| 296 |
+
"loss": 3.7122769165039062,
|
| 297 |
+
"step": 2050
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.01835007686825355,
|
| 301 |
+
"grad_norm": 0.2162218689918518,
|
| 302 |
+
"learning_rate": 0.001,
|
| 303 |
+
"loss": 3.690401611328125,
|
| 304 |
+
"step": 2100
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.018786983460354824,
|
| 308 |
+
"grad_norm": 0.2171812504529953,
|
| 309 |
+
"learning_rate": 0.001,
|
| 310 |
+
"loss": 3.6825027465820312,
|
| 311 |
+
"step": 2150
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.0192238900524561,
|
| 315 |
+
"grad_norm": 0.22677843272686005,
|
| 316 |
+
"learning_rate": 0.001,
|
| 317 |
+
"loss": 3.6679034423828125,
|
| 318 |
+
"step": 2200
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.019660796644557374,
|
| 322 |
+
"grad_norm": 0.23934350907802582,
|
| 323 |
+
"learning_rate": 0.001,
|
| 324 |
+
"loss": 3.6494000244140623,
|
| 325 |
+
"step": 2250
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.02009770323665865,
|
| 329 |
+
"grad_norm": 0.20819289982318878,
|
| 330 |
+
"learning_rate": 0.001,
|
| 331 |
+
"loss": 3.640907897949219,
|
| 332 |
+
"step": 2300
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.020534609828759924,
|
| 336 |
+
"grad_norm": 0.22890810668468475,
|
| 337 |
+
"learning_rate": 0.001,
|
| 338 |
+
"loss": 3.6309326171875,
|
| 339 |
+
"step": 2350
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.0209715164208612,
|
| 343 |
+
"grad_norm": 0.2085026353597641,
|
| 344 |
+
"learning_rate": 0.001,
|
| 345 |
+
"loss": 3.6261746215820314,
|
| 346 |
+
"step": 2400
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.021408423012962474,
|
| 350 |
+
"grad_norm": 0.22136099636554718,
|
| 351 |
+
"learning_rate": 0.001,
|
| 352 |
+
"loss": 3.614878845214844,
|
| 353 |
+
"step": 2450
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.021845329605063746,
|
| 357 |
+
"grad_norm": 0.21210630238056183,
|
| 358 |
+
"learning_rate": 0.001,
|
| 359 |
+
"loss": 3.599817810058594,
|
| 360 |
+
"step": 2500
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.02228223619716502,
|
| 364 |
+
"grad_norm": 0.24638274312019348,
|
| 365 |
+
"learning_rate": 0.001,
|
| 366 |
+
"loss": 3.591536865234375,
|
| 367 |
+
"step": 2550
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.022719142789266296,
|
| 371 |
+
"grad_norm": 0.20501184463500977,
|
| 372 |
+
"learning_rate": 0.001,
|
| 373 |
+
"loss": 3.58053955078125,
|
| 374 |
+
"step": 2600
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.02315604938136757,
|
| 378 |
+
"grad_norm": 0.2101815640926361,
|
| 379 |
+
"learning_rate": 0.001,
|
| 380 |
+
"loss": 3.5824459838867186,
|
| 381 |
+
"step": 2650
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.023592955973468846,
|
| 385 |
+
"grad_norm": 0.22906431555747986,
|
| 386 |
+
"learning_rate": 0.001,
|
| 387 |
+
"loss": 3.564821472167969,
|
| 388 |
+
"step": 2700
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.02402986256557012,
|
| 392 |
+
"grad_norm": 0.20708167552947998,
|
| 393 |
+
"learning_rate": 0.001,
|
| 394 |
+
"loss": 3.5566305541992187,
|
| 395 |
+
"step": 2750
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.024466769157671396,
|
| 399 |
+
"grad_norm": 0.21956811845302582,
|
| 400 |
+
"learning_rate": 0.001,
|
| 401 |
+
"loss": 3.5507748413085936,
|
| 402 |
+
"step": 2800
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.02490367574977267,
|
| 406 |
+
"grad_norm": 0.24124488234519958,
|
| 407 |
+
"learning_rate": 0.001,
|
| 408 |
+
"loss": 3.5426095581054686,
|
| 409 |
+
"step": 2850
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.025340582341873946,
|
| 413 |
+
"grad_norm": 0.20921742916107178,
|
| 414 |
+
"learning_rate": 0.001,
|
| 415 |
+
"loss": 3.531602783203125,
|
| 416 |
+
"step": 2900
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.02577748893397522,
|
| 420 |
+
"grad_norm": 0.19601280987262726,
|
| 421 |
+
"learning_rate": 0.001,
|
| 422 |
+
"loss": 3.5225894165039064,
|
| 423 |
+
"step": 2950
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.026214395526076496,
|
| 427 |
+
"grad_norm": 0.20243392884731293,
|
| 428 |
+
"learning_rate": 0.001,
|
| 429 |
+
"loss": 3.5183840942382814,
|
| 430 |
+
"step": 3000
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.026214395526076496,
|
| 434 |
+
"eval_loss": 3.518667221069336,
|
| 435 |
+
"eval_runtime": 5.8945,
|
| 436 |
+
"eval_samples_per_second": 165.579,
|
| 437 |
+
"eval_steps_per_second": 10.349,
|
| 438 |
+
"step": 3000
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 0.02665130211817777,
|
| 442 |
+
"grad_norm": 0.1983480155467987,
|
| 443 |
+
"learning_rate": 0.001,
|
| 444 |
+
"loss": 3.5175689697265624,
|
| 445 |
+
"step": 3050
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 0.027088208710279046,
|
| 449 |
+
"grad_norm": 0.20962044596672058,
|
| 450 |
+
"learning_rate": 0.001,
|
| 451 |
+
"loss": 3.501134033203125,
|
| 452 |
+
"step": 3100
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 0.02752511530238032,
|
| 456 |
+
"grad_norm": 0.20161673426628113,
|
| 457 |
+
"learning_rate": 0.001,
|
| 458 |
+
"loss": 3.508039855957031,
|
| 459 |
+
"step": 3150
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 0.027962021894481597,
|
| 463 |
+
"grad_norm": 0.21198803186416626,
|
| 464 |
+
"learning_rate": 0.001,
|
| 465 |
+
"loss": 3.493387451171875,
|
| 466 |
+
"step": 3200
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 0.02839892848658287,
|
| 470 |
+
"grad_norm": 0.21940743923187256,
|
| 471 |
+
"learning_rate": 0.001,
|
| 472 |
+
"loss": 3.4941644287109375,
|
| 473 |
+
"step": 3250
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 0.028835835078684147,
|
| 477 |
+
"grad_norm": 0.23343515396118164,
|
| 478 |
+
"learning_rate": 0.001,
|
| 479 |
+
"loss": 3.4835015869140626,
|
| 480 |
+
"step": 3300
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 0.02927274167078542,
|
| 484 |
+
"grad_norm": 0.19761498272418976,
|
| 485 |
+
"learning_rate": 0.001,
|
| 486 |
+
"loss": 3.482951965332031,
|
| 487 |
+
"step": 3350
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 0.029709648262886697,
|
| 491 |
+
"grad_norm": 0.22693106532096863,
|
| 492 |
+
"learning_rate": 0.001,
|
| 493 |
+
"loss": 3.475736389160156,
|
| 494 |
+
"step": 3400
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.030146554854987972,
|
| 498 |
+
"grad_norm": 0.19969327747821808,
|
| 499 |
+
"learning_rate": 0.001,
|
| 500 |
+
"loss": 3.4689141845703126,
|
| 501 |
+
"step": 3450
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 0.030583461447089247,
|
| 505 |
+
"grad_norm": 0.23987655341625214,
|
| 506 |
+
"learning_rate": 0.001,
|
| 507 |
+
"loss": 3.464185791015625,
|
| 508 |
+
"step": 3500
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 0.031020368039190522,
|
| 512 |
+
"grad_norm": 0.20894227921962738,
|
| 513 |
+
"learning_rate": 0.001,
|
| 514 |
+
"loss": 3.4571575927734375,
|
| 515 |
+
"step": 3550
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 0.0314572746312918,
|
| 519 |
+
"grad_norm": 0.22110916674137115,
|
| 520 |
+
"learning_rate": 0.001,
|
| 521 |
+
"loss": 3.4559121704101563,
|
| 522 |
+
"step": 3600
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 0.03189418122339307,
|
| 526 |
+
"grad_norm": 0.21295017004013062,
|
| 527 |
+
"learning_rate": 0.001,
|
| 528 |
+
"loss": 3.451357421875,
|
| 529 |
+
"step": 3650
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 0.03233108781549435,
|
| 533 |
+
"grad_norm": 0.19029361009597778,
|
| 534 |
+
"learning_rate": 0.001,
|
| 535 |
+
"loss": 3.4517440795898438,
|
| 536 |
+
"step": 3700
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 0.03276799440759562,
|
| 540 |
+
"grad_norm": 0.2307061105966568,
|
| 541 |
+
"learning_rate": 0.001,
|
| 542 |
+
"loss": 3.4454116821289062,
|
| 543 |
+
"step": 3750
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 0.0332049009996969,
|
| 547 |
+
"grad_norm": 0.2172439694404602,
|
| 548 |
+
"learning_rate": 0.001,
|
| 549 |
+
"loss": 3.4348458862304687,
|
| 550 |
+
"step": 3800
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 0.03364180759179817,
|
| 554 |
+
"grad_norm": 0.19329442083835602,
|
| 555 |
+
"learning_rate": 0.001,
|
| 556 |
+
"loss": 3.437733154296875,
|
| 557 |
+
"step": 3850
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 0.03407871418389945,
|
| 561 |
+
"grad_norm": 0.2241591513156891,
|
| 562 |
+
"learning_rate": 0.001,
|
| 563 |
+
"loss": 3.426455383300781,
|
| 564 |
+
"step": 3900
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 0.03451562077600072,
|
| 568 |
+
"grad_norm": 0.2319568693637848,
|
| 569 |
+
"learning_rate": 0.001,
|
| 570 |
+
"loss": 3.433238830566406,
|
| 571 |
+
"step": 3950
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 0.034952527368102,
|
| 575 |
+
"grad_norm": 0.1924877166748047,
|
| 576 |
+
"learning_rate": 0.001,
|
| 577 |
+
"loss": 3.4249606323242188,
|
| 578 |
+
"step": 4000
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 0.03538943396020327,
|
| 582 |
+
"grad_norm": 0.2120707631111145,
|
| 583 |
+
"learning_rate": 0.001,
|
| 584 |
+
"loss": 3.421075439453125,
|
| 585 |
+
"step": 4050
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 0.03582634055230455,
|
| 589 |
+
"grad_norm": 0.19918771088123322,
|
| 590 |
+
"learning_rate": 0.001,
|
| 591 |
+
"loss": 3.4203054809570315,
|
| 592 |
+
"step": 4100
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 0.03626324714440582,
|
| 596 |
+
"grad_norm": 0.2003307044506073,
|
| 597 |
+
"learning_rate": 0.001,
|
| 598 |
+
"loss": 3.411707763671875,
|
| 599 |
+
"step": 4150
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 0.0367001537365071,
|
| 603 |
+
"grad_norm": 0.2270653247833252,
|
| 604 |
+
"learning_rate": 0.001,
|
| 605 |
+
"loss": 3.4078097534179688,
|
| 606 |
+
"step": 4200
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 0.03713706032860837,
|
| 610 |
+
"grad_norm": 0.21652576327323914,
|
| 611 |
+
"learning_rate": 0.001,
|
| 612 |
+
"loss": 3.4062127685546875,
|
| 613 |
+
"step": 4250
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 0.03757396692070965,
|
| 617 |
+
"grad_norm": 0.23578432202339172,
|
| 618 |
+
"learning_rate": 0.001,
|
| 619 |
+
"loss": 3.400606689453125,
|
| 620 |
+
"step": 4300
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 0.03801087351281092,
|
| 624 |
+
"grad_norm": 0.20475627481937408,
|
| 625 |
+
"learning_rate": 0.001,
|
| 626 |
+
"loss": 3.395752258300781,
|
| 627 |
+
"step": 4350
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 0.0384477801049122,
|
| 631 |
+
"grad_norm": 0.2542702555656433,
|
| 632 |
+
"learning_rate": 0.001,
|
| 633 |
+
"loss": 3.396254577636719,
|
| 634 |
+
"step": 4400
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 0.03888468669701347,
|
| 638 |
+
"grad_norm": 0.18298493325710297,
|
| 639 |
+
"learning_rate": 0.001,
|
| 640 |
+
"loss": 3.393648986816406,
|
| 641 |
+
"step": 4450
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 0.03932159328911475,
|
| 645 |
+
"grad_norm": 0.22572273015975952,
|
| 646 |
+
"learning_rate": 0.001,
|
| 647 |
+
"loss": 3.3865216064453123,
|
| 648 |
+
"step": 4500
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 0.03975849988121602,
|
| 652 |
+
"grad_norm": 0.19484151899814606,
|
| 653 |
+
"learning_rate": 0.001,
|
| 654 |
+
"loss": 3.3849478149414063,
|
| 655 |
+
"step": 4550
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 0.0401954064733173,
|
| 659 |
+
"grad_norm": 0.19386635720729828,
|
| 660 |
+
"learning_rate": 0.001,
|
| 661 |
+
"loss": 3.38037109375,
|
| 662 |
+
"step": 4600
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 0.04063231306541857,
|
| 666 |
+
"grad_norm": 0.20124493539333344,
|
| 667 |
+
"learning_rate": 0.001,
|
| 668 |
+
"loss": 3.382728271484375,
|
| 669 |
+
"step": 4650
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 0.04106921965751985,
|
| 673 |
+
"grad_norm": 0.19963102042675018,
|
| 674 |
+
"learning_rate": 0.001,
|
| 675 |
+
"loss": 3.381324462890625,
|
| 676 |
+
"step": 4700
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 0.04150612624962112,
|
| 680 |
+
"grad_norm": 0.20611299574375153,
|
| 681 |
+
"learning_rate": 0.001,
|
| 682 |
+
"loss": 3.3788223266601562,
|
| 683 |
+
"step": 4750
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 0.0419430328417224,
|
| 687 |
+
"grad_norm": 0.19405238330364227,
|
| 688 |
+
"learning_rate": 0.001,
|
| 689 |
+
"loss": 3.3741259765625,
|
| 690 |
+
"step": 4800
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 0.04237993943382367,
|
| 694 |
+
"grad_norm": 0.20914608240127563,
|
| 695 |
+
"learning_rate": 0.001,
|
| 696 |
+
"loss": 3.370547180175781,
|
| 697 |
+
"step": 4850
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 0.04281684602592495,
|
| 701 |
+
"grad_norm": 0.19159191846847534,
|
| 702 |
+
"learning_rate": 0.001,
|
| 703 |
+
"loss": 3.3625115966796875,
|
| 704 |
+
"step": 4900
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 0.04325375261802622,
|
| 708 |
+
"grad_norm": 0.19163894653320312,
|
| 709 |
+
"learning_rate": 0.001,
|
| 710 |
+
"loss": 3.360504455566406,
|
| 711 |
+
"step": 4950
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 0.04369065921012749,
|
| 715 |
+
"grad_norm": 0.19763918220996857,
|
| 716 |
+
"learning_rate": 0.001,
|
| 717 |
+
"loss": 3.3601028442382814,
|
| 718 |
+
"step": 5000
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 0.04412756580222877,
|
| 722 |
+
"grad_norm": 0.18974512815475464,
|
| 723 |
+
"learning_rate": 0.001,
|
| 724 |
+
"loss": 3.3579150390625,
|
| 725 |
+
"step": 5050
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 0.04456447239433004,
|
| 729 |
+
"grad_norm": 0.21315094828605652,
|
| 730 |
+
"learning_rate": 0.001,
|
| 731 |
+
"loss": 3.355892333984375,
|
| 732 |
+
"step": 5100
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 0.04500137898643132,
|
| 736 |
+
"grad_norm": 0.1904531568288803,
|
| 737 |
+
"learning_rate": 0.001,
|
| 738 |
+
"loss": 3.360745849609375,
|
| 739 |
+
"step": 5150
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 0.04543828557853259,
|
| 743 |
+
"grad_norm": 0.19907866418361664,
|
| 744 |
+
"learning_rate": 0.001,
|
| 745 |
+
"loss": 3.357358703613281,
|
| 746 |
+
"step": 5200
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 0.04587519217063387,
|
| 750 |
+
"grad_norm": 0.22703590989112854,
|
| 751 |
+
"learning_rate": 0.001,
|
| 752 |
+
"loss": 3.356294860839844,
|
| 753 |
+
"step": 5250
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 0.04631209876273514,
|
| 757 |
+
"grad_norm": 0.1892520785331726,
|
| 758 |
+
"learning_rate": 0.001,
|
| 759 |
+
"loss": 3.3490374755859373,
|
| 760 |
+
"step": 5300
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 0.04674900535483642,
|
| 764 |
+
"grad_norm": 0.18953296542167664,
|
| 765 |
+
"learning_rate": 0.001,
|
| 766 |
+
"loss": 3.3442022705078127,
|
| 767 |
+
"step": 5350
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 0.04718591194693769,
|
| 771 |
+
"grad_norm": 0.1933782547712326,
|
| 772 |
+
"learning_rate": 0.001,
|
| 773 |
+
"loss": 3.3514590454101563,
|
| 774 |
+
"step": 5400
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 0.04762281853903897,
|
| 778 |
+
"grad_norm": 0.22573590278625488,
|
| 779 |
+
"learning_rate": 0.001,
|
| 780 |
+
"loss": 3.3468975830078125,
|
| 781 |
+
"step": 5450
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 0.04805972513114024,
|
| 785 |
+
"grad_norm": 0.2123604267835617,
|
| 786 |
+
"learning_rate": 0.001,
|
| 787 |
+
"loss": 3.3379788208007812,
|
| 788 |
+
"step": 5500
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 0.04849663172324152,
|
| 792 |
+
"grad_norm": 0.18217475712299347,
|
| 793 |
+
"learning_rate": 0.001,
|
| 794 |
+
"loss": 3.344074401855469,
|
| 795 |
+
"step": 5550
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 0.04893353831534279,
|
| 799 |
+
"grad_norm": 0.2128428965806961,
|
| 800 |
+
"learning_rate": 0.001,
|
| 801 |
+
"loss": 3.3349212646484374,
|
| 802 |
+
"step": 5600
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 0.04937044490744407,
|
| 806 |
+
"grad_norm": 0.19785532355308533,
|
| 807 |
+
"learning_rate": 0.001,
|
| 808 |
+
"loss": 3.327950744628906,
|
| 809 |
+
"step": 5650
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 0.04980735149954534,
|
| 813 |
+
"grad_norm": 0.18241338431835175,
|
| 814 |
+
"learning_rate": 0.001,
|
| 815 |
+
"loss": 3.334012451171875,
|
| 816 |
+
"step": 5700
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 0.05024425809164662,
|
| 820 |
+
"grad_norm": 0.2020372748374939,
|
| 821 |
+
"learning_rate": 0.001,
|
| 822 |
+
"loss": 3.3329043579101563,
|
| 823 |
+
"step": 5750
|
| 824 |
+
},
|
| 825 |
+
{
|
| 826 |
+
"epoch": 0.05068116468374789,
|
| 827 |
+
"grad_norm": 0.2071504294872284,
|
| 828 |
+
"learning_rate": 0.001,
|
| 829 |
+
"loss": 3.33412353515625,
|
| 830 |
+
"step": 5800
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 0.05111807127584917,
|
| 834 |
+
"grad_norm": 0.20365303754806519,
|
| 835 |
+
"learning_rate": 0.001,
|
| 836 |
+
"loss": 3.322493591308594,
|
| 837 |
+
"step": 5850
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 0.05155497786795044,
|
| 841 |
+
"grad_norm": 0.21742033958435059,
|
| 842 |
+
"learning_rate": 0.001,
|
| 843 |
+
"loss": 3.329978332519531,
|
| 844 |
+
"step": 5900
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 0.05199188446005172,
|
| 848 |
+
"grad_norm": 0.18907567858695984,
|
| 849 |
+
"learning_rate": 0.001,
|
| 850 |
+
"loss": 3.328219299316406,
|
| 851 |
+
"step": 5950
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"epoch": 0.05242879105215299,
|
| 855 |
+
"grad_norm": 0.18905521929264069,
|
| 856 |
+
"learning_rate": 0.001,
|
| 857 |
+
"loss": 3.323075866699219,
|
| 858 |
+
"step": 6000
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 0.05242879105215299,
|
| 862 |
+
"eval_loss": 3.3261899948120117,
|
| 863 |
+
"eval_runtime": 6.1096,
|
| 864 |
+
"eval_samples_per_second": 159.748,
|
| 865 |
+
"eval_steps_per_second": 9.984,
|
| 866 |
+
"step": 6000
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 0.05286569764425427,
|
| 870 |
+
"grad_norm": 0.19901056587696075,
|
| 871 |
+
"learning_rate": 0.001,
|
| 872 |
+
"loss": 3.3208355712890625,
|
| 873 |
+
"step": 6050
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 0.05330260423635554,
|
| 877 |
+
"grad_norm": 0.19175691902637482,
|
| 878 |
+
"learning_rate": 0.001,
|
| 879 |
+
"loss": 3.3096588134765623,
|
| 880 |
+
"step": 6100
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"epoch": 0.05373951082845682,
|
| 884 |
+
"grad_norm": 0.20367567241191864,
|
| 885 |
+
"learning_rate": 0.001,
|
| 886 |
+
"loss": 3.318222961425781,
|
| 887 |
+
"step": 6150
|
| 888 |
+
},
|
| 889 |
+
{
|
| 890 |
+
"epoch": 0.05417641742055809,
|
| 891 |
+
"grad_norm": 0.20697498321533203,
|
| 892 |
+
"learning_rate": 0.001,
|
| 893 |
+
"loss": 3.314552917480469,
|
| 894 |
+
"step": 6200
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 0.05461332401265937,
|
| 898 |
+
"grad_norm": 0.23248012363910675,
|
| 899 |
+
"learning_rate": 0.001,
|
| 900 |
+
"loss": 3.312541809082031,
|
| 901 |
+
"step": 6250
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 0.05505023060476064,
|
| 905 |
+
"grad_norm": 0.21030908823013306,
|
| 906 |
+
"learning_rate": 0.001,
|
| 907 |
+
"loss": 3.3117620849609377,
|
| 908 |
+
"step": 6300
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 0.05548713719686192,
|
| 912 |
+
"grad_norm": 0.19639821350574493,
|
| 913 |
+
"learning_rate": 0.001,
|
| 914 |
+
"loss": 3.314966125488281,
|
| 915 |
+
"step": 6350
|
| 916 |
+
},
|
| 917 |
+
{
|
| 918 |
+
"epoch": 0.05592404378896319,
|
| 919 |
+
"grad_norm": 0.1988273561000824,
|
| 920 |
+
"learning_rate": 0.001,
|
| 921 |
+
"loss": 3.308790588378906,
|
| 922 |
+
"step": 6400
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"epoch": 0.05636095038106447,
|
| 926 |
+
"grad_norm": 0.20665325224399567,
|
| 927 |
+
"learning_rate": 0.001,
|
| 928 |
+
"loss": 3.3096978759765623,
|
| 929 |
+
"step": 6450
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 0.05679785697316574,
|
| 933 |
+
"grad_norm": 0.1974414736032486,
|
| 934 |
+
"learning_rate": 0.001,
|
| 935 |
+
"loss": 3.3091079711914064,
|
| 936 |
+
"step": 6500
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"epoch": 0.057234763565267015,
|
| 940 |
+
"grad_norm": 0.19521687924861908,
|
| 941 |
+
"learning_rate": 0.001,
|
| 942 |
+
"loss": 3.3086837768554687,
|
| 943 |
+
"step": 6550
|
| 944 |
+
},
|
| 945 |
+
{
|
| 946 |
+
"epoch": 0.05767167015736829,
|
| 947 |
+
"grad_norm": 0.19938398897647858,
|
| 948 |
+
"learning_rate": 0.001,
|
| 949 |
+
"loss": 3.3038284301757814,
|
| 950 |
+
"step": 6600
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 0.058108576749469565,
|
| 954 |
+
"grad_norm": 0.18721790611743927,
|
| 955 |
+
"learning_rate": 0.001,
|
| 956 |
+
"loss": 3.3031985473632814,
|
| 957 |
+
"step": 6650
|
| 958 |
+
},
|
| 959 |
+
{
|
| 960 |
+
"epoch": 0.05854548334157084,
|
| 961 |
+
"grad_norm": 0.20039626955986023,
|
| 962 |
+
"learning_rate": 0.001,
|
| 963 |
+
"loss": 3.3011334228515623,
|
| 964 |
+
"step": 6700
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 0.058982389933672115,
|
| 968 |
+
"grad_norm": 0.22351515293121338,
|
| 969 |
+
"learning_rate": 0.001,
|
| 970 |
+
"loss": 3.3007638549804685,
|
| 971 |
+
"step": 6750
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 0.059419296525773393,
|
| 975 |
+
"grad_norm": 0.20300878584384918,
|
| 976 |
+
"learning_rate": 0.001,
|
| 977 |
+
"loss": 3.2940847778320315,
|
| 978 |
+
"step": 6800
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 0.059856203117874665,
|
| 982 |
+
"grad_norm": 0.20710642635822296,
|
| 983 |
+
"learning_rate": 0.001,
|
| 984 |
+
"loss": 3.297225036621094,
|
| 985 |
+
"step": 6850
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 0.060293109709975944,
|
| 989 |
+
"grad_norm": 0.1960466355085373,
|
| 990 |
+
"learning_rate": 0.001,
|
| 991 |
+
"loss": 3.2912692260742187,
|
| 992 |
+
"step": 6900
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 0.060730016302077215,
|
| 996 |
+
"grad_norm": 0.1825372278690338,
|
| 997 |
+
"learning_rate": 0.001,
|
| 998 |
+
"loss": 3.2928152465820313,
|
| 999 |
+
"step": 6950
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 0.061166922894178494,
|
| 1003 |
+
"grad_norm": 0.19677802920341492,
|
| 1004 |
+
"learning_rate": 0.001,
|
| 1005 |
+
"loss": 3.2997564697265624,
|
| 1006 |
+
"step": 7000
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 0.061603829486279765,
|
| 1010 |
+
"grad_norm": 0.21398645639419556,
|
| 1011 |
+
"learning_rate": 0.001,
|
| 1012 |
+
"loss": 3.293714904785156,
|
| 1013 |
+
"step": 7050
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 0.062040736078381044,
|
| 1017 |
+
"grad_norm": 0.20046885311603546,
|
| 1018 |
+
"learning_rate": 0.001,
|
| 1019 |
+
"loss": 3.2939657592773437,
|
| 1020 |
+
"step": 7100
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 0.062477642670482315,
|
| 1024 |
+
"grad_norm": 0.1991434544324875,
|
| 1025 |
+
"learning_rate": 0.001,
|
| 1026 |
+
"loss": 3.284585266113281,
|
| 1027 |
+
"step": 7150
|
| 1028 |
+
},
|
| 1029 |
+
{
|
| 1030 |
+
"epoch": 0.0629145492625836,
|
| 1031 |
+
"grad_norm": 0.20454499125480652,
|
| 1032 |
+
"learning_rate": 0.001,
|
| 1033 |
+
"loss": 3.2882635498046877,
|
| 1034 |
+
"step": 7200
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 0.06335145585468487,
|
| 1038 |
+
"grad_norm": 0.20718543231487274,
|
| 1039 |
+
"learning_rate": 0.001,
|
| 1040 |
+
"loss": 3.283497619628906,
|
| 1041 |
+
"step": 7250
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 0.06378836244678614,
|
| 1045 |
+
"grad_norm": 0.1971973031759262,
|
| 1046 |
+
"learning_rate": 0.001,
|
| 1047 |
+
"loss": 3.284901123046875,
|
| 1048 |
+
"step": 7300
|
| 1049 |
+
},
|
| 1050 |
+
{
|
| 1051 |
+
"epoch": 0.06422526903888742,
|
| 1052 |
+
"grad_norm": 0.18264292180538177,
|
| 1053 |
+
"learning_rate": 0.001,
|
| 1054 |
+
"loss": 3.285682373046875,
|
| 1055 |
+
"step": 7350
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 0.0646621756309887,
|
| 1059 |
+
"grad_norm": 0.2100597620010376,
|
| 1060 |
+
"learning_rate": 0.001,
|
| 1061 |
+
"loss": 3.284263000488281,
|
| 1062 |
+
"step": 7400
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 0.06509908222308997,
|
| 1066 |
+
"grad_norm": 0.19299328327178955,
|
| 1067 |
+
"learning_rate": 0.001,
|
| 1068 |
+
"loss": 3.283685302734375,
|
| 1069 |
+
"step": 7450
|
| 1070 |
+
},
|
| 1071 |
+
{
|
| 1072 |
+
"epoch": 0.06553598881519124,
|
| 1073 |
+
"grad_norm": 0.18712171912193298,
|
| 1074 |
+
"learning_rate": 0.001,
|
| 1075 |
+
"loss": 3.279105224609375,
|
| 1076 |
+
"step": 7500
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 0.06597289540729252,
|
| 1080 |
+
"grad_norm": 0.18977321684360504,
|
| 1081 |
+
"learning_rate": 0.001,
|
| 1082 |
+
"loss": 3.287223205566406,
|
| 1083 |
+
"step": 7550
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 0.0664098019993938,
|
| 1087 |
+
"grad_norm": 0.22269035875797272,
|
| 1088 |
+
"learning_rate": 0.001,
|
| 1089 |
+
"loss": 3.2770404052734374,
|
| 1090 |
+
"step": 7600
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 0.06684670859149507,
|
| 1094 |
+
"grad_norm": 0.18847301602363586,
|
| 1095 |
+
"learning_rate": 0.001,
|
| 1096 |
+
"loss": 3.2692807006835936,
|
| 1097 |
+
"step": 7650
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 0.06728361518359634,
|
| 1101 |
+
"grad_norm": 0.21769355237483978,
|
| 1102 |
+
"learning_rate": 0.001,
|
| 1103 |
+
"loss": 3.2741900634765626,
|
| 1104 |
+
"step": 7700
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 0.06772052177569762,
|
| 1108 |
+
"grad_norm": 0.18970917165279388,
|
| 1109 |
+
"learning_rate": 0.001,
|
| 1110 |
+
"loss": 3.2736630249023437,
|
| 1111 |
+
"step": 7750
|
| 1112 |
+
},
|
| 1113 |
+
{
|
| 1114 |
+
"epoch": 0.0681574283677989,
|
| 1115 |
+
"grad_norm": 0.19735926389694214,
|
| 1116 |
+
"learning_rate": 0.001,
|
| 1117 |
+
"loss": 3.27821044921875,
|
| 1118 |
+
"step": 7800
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 0.06859433495990017,
|
| 1122 |
+
"grad_norm": 0.19194583594799042,
|
| 1123 |
+
"learning_rate": 0.001,
|
| 1124 |
+
"loss": 3.2705389404296876,
|
| 1125 |
+
"step": 7850
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 0.06903124155200144,
|
| 1129 |
+
"grad_norm": 0.2058054655790329,
|
| 1130 |
+
"learning_rate": 0.001,
|
| 1131 |
+
"loss": 3.2689944458007814,
|
| 1132 |
+
"step": 7900
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 0.06946814814410272,
|
| 1136 |
+
"grad_norm": 0.18322297930717468,
|
| 1137 |
+
"learning_rate": 0.001,
|
| 1138 |
+
"loss": 3.2653887939453123,
|
| 1139 |
+
"step": 7950
|
| 1140 |
+
},
|
| 1141 |
+
{
|
| 1142 |
+
"epoch": 0.069905054736204,
|
| 1143 |
+
"grad_norm": 0.1952529400587082,
|
| 1144 |
+
"learning_rate": 0.001,
|
| 1145 |
+
"loss": 3.2640380859375,
|
| 1146 |
+
"step": 8000
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 0.07034196132830527,
|
| 1150 |
+
"grad_norm": 0.1916273683309555,
|
| 1151 |
+
"learning_rate": 0.001,
|
| 1152 |
+
"loss": 3.274902648925781,
|
| 1153 |
+
"step": 8050
|
| 1154 |
+
},
|
| 1155 |
+
{
|
| 1156 |
+
"epoch": 0.07077886792040654,
|
| 1157 |
+
"grad_norm": 0.21100108325481415,
|
| 1158 |
+
"learning_rate": 0.001,
|
| 1159 |
+
"loss": 3.265535583496094,
|
| 1160 |
+
"step": 8100
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 0.07121577451250782,
|
| 1164 |
+
"grad_norm": 0.2012566179037094,
|
| 1165 |
+
"learning_rate": 0.001,
|
| 1166 |
+
"loss": 3.2651483154296876,
|
| 1167 |
+
"step": 8150
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 0.0716526811046091,
|
| 1171 |
+
"grad_norm": 0.1913347840309143,
|
| 1172 |
+
"learning_rate": 0.001,
|
| 1173 |
+
"loss": 3.2659671020507814,
|
| 1174 |
+
"step": 8200
|
| 1175 |
+
},
|
| 1176 |
+
{
|
| 1177 |
+
"epoch": 0.07208958769671037,
|
| 1178 |
+
"grad_norm": 0.19437524676322937,
|
| 1179 |
+
"learning_rate": 0.001,
|
| 1180 |
+
"loss": 3.268343811035156,
|
| 1181 |
+
"step": 8250
|
| 1182 |
+
},
|
| 1183 |
+
{
|
| 1184 |
+
"epoch": 0.07252649428881164,
|
| 1185 |
+
"grad_norm": 0.20302678644657135,
|
| 1186 |
+
"learning_rate": 0.001,
|
| 1187 |
+
"loss": 3.270460510253906,
|
| 1188 |
+
"step": 8300
|
| 1189 |
+
},
|
| 1190 |
+
{
|
| 1191 |
+
"epoch": 0.07296340088091291,
|
| 1192 |
+
"grad_norm": 0.1874990612268448,
|
| 1193 |
+
"learning_rate": 0.001,
|
| 1194 |
+
"loss": 3.2629928588867188,
|
| 1195 |
+
"step": 8350
|
| 1196 |
+
},
|
| 1197 |
+
{
|
| 1198 |
+
"epoch": 0.0734003074730142,
|
| 1199 |
+
"grad_norm": 0.19520102441310883,
|
| 1200 |
+
"learning_rate": 0.001,
|
| 1201 |
+
"loss": 3.265036315917969,
|
| 1202 |
+
"step": 8400
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 0.07383721406511547,
|
| 1206 |
+
"grad_norm": 0.18731839954853058,
|
| 1207 |
+
"learning_rate": 0.001,
|
| 1208 |
+
"loss": 3.2578021240234376,
|
| 1209 |
+
"step": 8450
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 0.07427412065721674,
|
| 1213 |
+
"grad_norm": 0.20026656985282898,
|
| 1214 |
+
"learning_rate": 0.001,
|
| 1215 |
+
"loss": 3.2632598876953125,
|
| 1216 |
+
"step": 8500
|
| 1217 |
+
},
|
| 1218 |
+
{
|
| 1219 |
+
"epoch": 0.07471102724931801,
|
| 1220 |
+
"grad_norm": 0.20248965919017792,
|
| 1221 |
+
"learning_rate": 0.001,
|
| 1222 |
+
"loss": 3.260670471191406,
|
| 1223 |
+
"step": 8550
|
| 1224 |
+
},
|
| 1225 |
+
{
|
| 1226 |
+
"epoch": 0.0751479338414193,
|
| 1227 |
+
"grad_norm": 0.20652979612350464,
|
| 1228 |
+
"learning_rate": 0.001,
|
| 1229 |
+
"loss": 3.2585479736328127,
|
| 1230 |
+
"step": 8600
|
| 1231 |
+
},
|
| 1232 |
+
{
|
| 1233 |
+
"epoch": 0.07558484043352057,
|
| 1234 |
+
"grad_norm": 0.18648388981819153,
|
| 1235 |
+
"learning_rate": 0.001,
|
| 1236 |
+
"loss": 3.259482727050781,
|
| 1237 |
+
"step": 8650
|
| 1238 |
+
},
|
| 1239 |
+
{
|
| 1240 |
+
"epoch": 0.07602174702562184,
|
| 1241 |
+
"grad_norm": 0.20123572647571564,
|
| 1242 |
+
"learning_rate": 0.001,
|
| 1243 |
+
"loss": 3.2517401123046876,
|
| 1244 |
+
"step": 8700
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 0.07645865361772311,
|
| 1248 |
+
"grad_norm": 0.20070253312587738,
|
| 1249 |
+
"learning_rate": 0.001,
|
| 1250 |
+
"loss": 3.2553228759765624,
|
| 1251 |
+
"step": 8750
|
| 1252 |
+
},
|
| 1253 |
+
{
|
| 1254 |
+
"epoch": 0.0768955602098244,
|
| 1255 |
+
"grad_norm": 0.19734439253807068,
|
| 1256 |
+
"learning_rate": 0.001,
|
| 1257 |
+
"loss": 3.2518017578125,
|
| 1258 |
+
"step": 8800
|
| 1259 |
+
},
|
| 1260 |
+
{
|
| 1261 |
+
"epoch": 0.07733246680192567,
|
| 1262 |
+
"grad_norm": 0.2070627510547638,
|
| 1263 |
+
"learning_rate": 0.001,
|
| 1264 |
+
"loss": 3.251908874511719,
|
| 1265 |
+
"step": 8850
|
| 1266 |
+
},
|
| 1267 |
+
{
|
| 1268 |
+
"epoch": 0.07776937339402694,
|
| 1269 |
+
"grad_norm": 0.19075073301792145,
|
| 1270 |
+
"learning_rate": 0.001,
|
| 1271 |
+
"loss": 3.2540167236328124,
|
| 1272 |
+
"step": 8900
|
| 1273 |
+
},
|
| 1274 |
+
{
|
| 1275 |
+
"epoch": 0.07820627998612821,
|
| 1276 |
+
"grad_norm": 0.2121705412864685,
|
| 1277 |
+
"learning_rate": 0.001,
|
| 1278 |
+
"loss": 3.253029479980469,
|
| 1279 |
+
"step": 8950
|
| 1280 |
+
},
|
| 1281 |
+
{
|
| 1282 |
+
"epoch": 0.0786431865782295,
|
| 1283 |
+
"grad_norm": 0.21024636924266815,
|
| 1284 |
+
"learning_rate": 0.001,
|
| 1285 |
+
"loss": 3.2557443237304686,
|
| 1286 |
+
"step": 9000
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 0.0786431865782295,
|
| 1290 |
+
"eval_loss": 3.257852554321289,
|
| 1291 |
+
"eval_runtime": 6.3455,
|
| 1292 |
+
"eval_samples_per_second": 153.811,
|
| 1293 |
+
"eval_steps_per_second": 9.613,
|
| 1294 |
+
"step": 9000
|
| 1295 |
+
},
|
| 1296 |
+
{
|
| 1297 |
+
"epoch": 0.07908009317033077,
|
| 1298 |
+
"grad_norm": 0.18182849884033203,
|
| 1299 |
+
"learning_rate": 0.001,
|
| 1300 |
+
"loss": 3.255887451171875,
|
| 1301 |
+
"step": 9050
|
| 1302 |
+
},
|
| 1303 |
+
{
|
| 1304 |
+
"epoch": 0.07951699976243204,
|
| 1305 |
+
"grad_norm": 0.20949381589889526,
|
| 1306 |
+
"learning_rate": 0.001,
|
| 1307 |
+
"loss": 3.2509515380859373,
|
| 1308 |
+
"step": 9100
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 0.07995390635453331,
|
| 1312 |
+
"grad_norm": 0.1916966736316681,
|
| 1313 |
+
"learning_rate": 0.001,
|
| 1314 |
+
"loss": 3.2475936889648436,
|
| 1315 |
+
"step": 9150
|
| 1316 |
+
},
|
| 1317 |
+
{
|
| 1318 |
+
"epoch": 0.0803908129466346,
|
| 1319 |
+
"grad_norm": 0.195968896150589,
|
| 1320 |
+
"learning_rate": 0.001,
|
| 1321 |
+
"loss": 3.2473486328125,
|
| 1322 |
+
"step": 9200
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 0.08082771953873587,
|
| 1326 |
+
"grad_norm": 0.19571411609649658,
|
| 1327 |
+
"learning_rate": 0.001,
|
| 1328 |
+
"loss": 3.2515054321289063,
|
| 1329 |
+
"step": 9250
|
| 1330 |
+
},
|
| 1331 |
+
{
|
| 1332 |
+
"epoch": 0.08126462613083714,
|
| 1333 |
+
"grad_norm": 0.1898670643568039,
|
| 1334 |
+
"learning_rate": 0.001,
|
| 1335 |
+
"loss": 3.248476257324219,
|
| 1336 |
+
"step": 9300
|
| 1337 |
+
},
|
| 1338 |
+
{
|
| 1339 |
+
"epoch": 0.08170153272293841,
|
| 1340 |
+
"grad_norm": 0.1957128793001175,
|
| 1341 |
+
"learning_rate": 0.001,
|
| 1342 |
+
"loss": 3.2485791015625,
|
| 1343 |
+
"step": 9350
|
| 1344 |
+
},
|
| 1345 |
+
{
|
| 1346 |
+
"epoch": 0.0821384393150397,
|
| 1347 |
+
"grad_norm": 0.1989963799715042,
|
| 1348 |
+
"learning_rate": 0.001,
|
| 1349 |
+
"loss": 3.2457614135742188,
|
| 1350 |
+
"step": 9400
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 0.08257534590714097,
|
| 1354 |
+
"grad_norm": 0.18394291400909424,
|
| 1355 |
+
"learning_rate": 0.001,
|
| 1356 |
+
"loss": 3.245361328125,
|
| 1357 |
+
"step": 9450
|
| 1358 |
+
},
|
| 1359 |
+
{
|
| 1360 |
+
"epoch": 0.08301225249924224,
|
| 1361 |
+
"grad_norm": 0.17211784422397614,
|
| 1362 |
+
"learning_rate": 0.001,
|
| 1363 |
+
"loss": 3.2452508544921876,
|
| 1364 |
+
"step": 9500
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 0.08344915909134351,
|
| 1368 |
+
"grad_norm": 0.20545506477355957,
|
| 1369 |
+
"learning_rate": 0.001,
|
| 1370 |
+
"loss": 3.2405926513671877,
|
| 1371 |
+
"step": 9550
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 0.0838860656834448,
|
| 1375 |
+
"grad_norm": 0.20575234293937683,
|
| 1376 |
+
"learning_rate": 0.001,
|
| 1377 |
+
"loss": 3.2421484375,
|
| 1378 |
+
"step": 9600
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 0.08432297227554607,
|
| 1382 |
+
"grad_norm": 0.19145144522190094,
|
| 1383 |
+
"learning_rate": 0.001,
|
| 1384 |
+
"loss": 3.2430648803710938,
|
| 1385 |
+
"step": 9650
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 0.08475987886764734,
|
| 1389 |
+
"grad_norm": 0.20513130724430084,
|
| 1390 |
+
"learning_rate": 0.001,
|
| 1391 |
+
"loss": 3.241943664550781,
|
| 1392 |
+
"step": 9700
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 0.08519678545974861,
|
| 1396 |
+
"grad_norm": 0.18931901454925537,
|
| 1397 |
+
"learning_rate": 0.001,
|
| 1398 |
+
"loss": 3.242476806640625,
|
| 1399 |
+
"step": 9750
|
| 1400 |
+
},
|
| 1401 |
+
{
|
| 1402 |
+
"epoch": 0.0856336920518499,
|
| 1403 |
+
"grad_norm": 0.2036847621202469,
|
| 1404 |
+
"learning_rate": 0.001,
|
| 1405 |
+
"loss": 3.2381185913085937,
|
| 1406 |
+
"step": 9800
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 0.08607059864395117,
|
| 1410 |
+
"grad_norm": 0.18806156516075134,
|
| 1411 |
+
"learning_rate": 0.001,
|
| 1412 |
+
"loss": 3.2394903564453124,
|
| 1413 |
+
"step": 9850
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 0.08650750523605244,
|
| 1417 |
+
"grad_norm": 0.18992380797863007,
|
| 1418 |
+
"learning_rate": 0.001,
|
| 1419 |
+
"loss": 3.2361404418945314,
|
| 1420 |
+
"step": 9900
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 0.08694441182815371,
|
| 1424 |
+
"grad_norm": 0.1844012290239334,
|
| 1425 |
+
"learning_rate": 0.001,
|
| 1426 |
+
"loss": 3.238492126464844,
|
| 1427 |
+
"step": 9950
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 0.08738131842025498,
|
| 1431 |
+
"grad_norm": 0.19030481576919556,
|
| 1432 |
+
"learning_rate": 0.001,
|
| 1433 |
+
"loss": 3.2360128784179687,
|
| 1434 |
+
"step": 10000
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 0.08781822501235627,
|
| 1438 |
+
"grad_norm": 0.20944413542747498,
|
| 1439 |
+
"learning_rate": 0.001,
|
| 1440 |
+
"loss": 3.23896728515625,
|
| 1441 |
+
"step": 10050
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 0.08825513160445754,
|
| 1445 |
+
"grad_norm": 0.1730359047651291,
|
| 1446 |
+
"learning_rate": 0.001,
|
| 1447 |
+
"loss": 3.233212890625,
|
| 1448 |
+
"step": 10100
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 0.08869203819655881,
|
| 1452 |
+
"grad_norm": 0.18298979103565216,
|
| 1453 |
+
"learning_rate": 0.001,
|
| 1454 |
+
"loss": 3.2336996459960936,
|
| 1455 |
+
"step": 10150
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"epoch": 0.08912894478866008,
|
| 1459 |
+
"grad_norm": 0.1836487203836441,
|
| 1460 |
+
"learning_rate": 0.001,
|
| 1461 |
+
"loss": 3.239130859375,
|
| 1462 |
+
"step": 10200
|
| 1463 |
+
},
|
| 1464 |
+
{
|
| 1465 |
+
"epoch": 0.08956585138076137,
|
| 1466 |
+
"grad_norm": 0.20185187458992004,
|
| 1467 |
+
"learning_rate": 0.001,
|
| 1468 |
+
"loss": 3.2354965209960938,
|
| 1469 |
+
"step": 10250
|
| 1470 |
+
},
|
| 1471 |
+
{
|
| 1472 |
+
"epoch": 0.09000275797286264,
|
| 1473 |
+
"grad_norm": 0.20701956748962402,
|
| 1474 |
+
"learning_rate": 0.001,
|
| 1475 |
+
"loss": 3.2316860961914062,
|
| 1476 |
+
"step": 10300
|
| 1477 |
+
},
|
| 1478 |
+
{
|
| 1479 |
+
"epoch": 0.09043966456496391,
|
| 1480 |
+
"grad_norm": 0.17721368372440338,
|
| 1481 |
+
"learning_rate": 0.001,
|
| 1482 |
+
"loss": 3.2350833129882814,
|
| 1483 |
+
"step": 10350
|
| 1484 |
+
},
|
| 1485 |
+
{
|
| 1486 |
+
"epoch": 0.09087657115706518,
|
| 1487 |
+
"grad_norm": 0.1922338306903839,
|
| 1488 |
+
"learning_rate": 0.001,
|
| 1489 |
+
"loss": 3.2318270874023436,
|
| 1490 |
+
"step": 10400
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"epoch": 0.09131347774916647,
|
| 1494 |
+
"grad_norm": 0.1857219636440277,
|
| 1495 |
+
"learning_rate": 0.001,
|
| 1496 |
+
"loss": 3.2357696533203124,
|
| 1497 |
+
"step": 10450
|
| 1498 |
+
},
|
| 1499 |
+
{
|
| 1500 |
+
"epoch": 0.09175038434126774,
|
| 1501 |
+
"grad_norm": 0.18690387904644012,
|
| 1502 |
+
"learning_rate": 0.001,
|
| 1503 |
+
"loss": 3.233714599609375,
|
| 1504 |
+
"step": 10500
|
| 1505 |
+
},
|
| 1506 |
+
{
|
| 1507 |
+
"epoch": 0.09218729093336901,
|
| 1508 |
+
"grad_norm": 0.22552669048309326,
|
| 1509 |
+
"learning_rate": 0.001,
|
| 1510 |
+
"loss": 3.2236834716796876,
|
| 1511 |
+
"step": 10550
|
| 1512 |
+
},
|
| 1513 |
+
{
|
| 1514 |
+
"epoch": 0.09262419752547028,
|
| 1515 |
+
"grad_norm": 0.2104293406009674,
|
| 1516 |
+
"learning_rate": 0.001,
|
| 1517 |
+
"loss": 3.2278411865234373,
|
| 1518 |
+
"step": 10600
|
| 1519 |
+
},
|
| 1520 |
+
{
|
| 1521 |
+
"epoch": 0.09306110411757157,
|
| 1522 |
+
"grad_norm": 0.1924365907907486,
|
| 1523 |
+
"learning_rate": 0.001,
|
| 1524 |
+
"loss": 3.2275039672851564,
|
| 1525 |
+
"step": 10650
|
| 1526 |
+
},
|
| 1527 |
+
{
|
| 1528 |
+
"epoch": 0.09349801070967284,
|
| 1529 |
+
"grad_norm": 0.17614150047302246,
|
| 1530 |
+
"learning_rate": 0.001,
|
| 1531 |
+
"loss": 3.2285580444335937,
|
| 1532 |
+
"step": 10700
|
| 1533 |
+
},
|
| 1534 |
+
{
|
| 1535 |
+
"epoch": 0.09393491730177411,
|
| 1536 |
+
"grad_norm": 0.1965877264738083,
|
| 1537 |
+
"learning_rate": 0.001,
|
| 1538 |
+
"loss": 3.23059814453125,
|
| 1539 |
+
"step": 10750
|
| 1540 |
+
},
|
| 1541 |
+
{
|
| 1542 |
+
"epoch": 0.09437182389387538,
|
| 1543 |
+
"grad_norm": 0.17845283448696136,
|
| 1544 |
+
"learning_rate": 0.001,
|
| 1545 |
+
"loss": 3.225852355957031,
|
| 1546 |
+
"step": 10800
|
| 1547 |
+
},
|
| 1548 |
+
{
|
| 1549 |
+
"epoch": 0.09480873048597667,
|
| 1550 |
+
"grad_norm": 0.1709090620279312,
|
| 1551 |
+
"learning_rate": 0.001,
|
| 1552 |
+
"loss": 3.2312142944335935,
|
| 1553 |
+
"step": 10850
|
| 1554 |
+
},
|
| 1555 |
+
{
|
| 1556 |
+
"epoch": 0.09524563707807794,
|
| 1557 |
+
"grad_norm": 0.17048615217208862,
|
| 1558 |
+
"learning_rate": 0.001,
|
| 1559 |
+
"loss": 3.22577392578125,
|
| 1560 |
+
"step": 10900
|
| 1561 |
+
},
|
| 1562 |
+
{
|
| 1563 |
+
"epoch": 0.09568254367017921,
|
| 1564 |
+
"grad_norm": 0.2272968739271164,
|
| 1565 |
+
"learning_rate": 0.001,
|
| 1566 |
+
"loss": 3.2276971435546873,
|
| 1567 |
+
"step": 10950
|
| 1568 |
+
},
|
| 1569 |
+
{
|
| 1570 |
+
"epoch": 0.09611945026228048,
|
| 1571 |
+
"grad_norm": 0.19360628724098206,
|
| 1572 |
+
"learning_rate": 0.001,
|
| 1573 |
+
"loss": 3.2247369384765623,
|
| 1574 |
+
"step": 11000
|
| 1575 |
+
},
|
| 1576 |
+
{
|
| 1577 |
+
"epoch": 0.09655635685438177,
|
| 1578 |
+
"grad_norm": 0.20023608207702637,
|
| 1579 |
+
"learning_rate": 0.001,
|
| 1580 |
+
"loss": 3.2287957763671873,
|
| 1581 |
+
"step": 11050
|
| 1582 |
+
},
|
| 1583 |
+
{
|
| 1584 |
+
"epoch": 0.09699326344648304,
|
| 1585 |
+
"grad_norm": 0.19376908242702484,
|
| 1586 |
+
"learning_rate": 0.001,
|
| 1587 |
+
"loss": 3.2265338134765624,
|
| 1588 |
+
"step": 11100
|
| 1589 |
+
},
|
| 1590 |
+
{
|
| 1591 |
+
"epoch": 0.09743017003858431,
|
| 1592 |
+
"grad_norm": 0.196820467710495,
|
| 1593 |
+
"learning_rate": 0.001,
|
| 1594 |
+
"loss": 3.218832702636719,
|
| 1595 |
+
"step": 11150
|
| 1596 |
+
},
|
| 1597 |
+
{
|
| 1598 |
+
"epoch": 0.09786707663068558,
|
| 1599 |
+
"grad_norm": 0.1909375935792923,
|
| 1600 |
+
"learning_rate": 0.001,
|
| 1601 |
+
"loss": 3.2203463745117187,
|
| 1602 |
+
"step": 11200
|
| 1603 |
+
},
|
| 1604 |
+
{
|
| 1605 |
+
"epoch": 0.09830398322278687,
|
| 1606 |
+
"grad_norm": 0.18288679420948029,
|
| 1607 |
+
"learning_rate": 0.001,
|
| 1608 |
+
"loss": 3.218720703125,
|
| 1609 |
+
"step": 11250
|
| 1610 |
+
},
|
| 1611 |
+
{
|
| 1612 |
+
"epoch": 0.09874088981488814,
|
| 1613 |
+
"grad_norm": 0.18179824948310852,
|
| 1614 |
+
"learning_rate": 0.001,
|
| 1615 |
+
"loss": 3.217382507324219,
|
| 1616 |
+
"step": 11300
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 0.09917779640698941,
|
| 1620 |
+
"grad_norm": 0.18757817149162292,
|
| 1621 |
+
"learning_rate": 0.001,
|
| 1622 |
+
"loss": 3.222406311035156,
|
| 1623 |
+
"step": 11350
|
| 1624 |
+
},
|
| 1625 |
+
{
|
| 1626 |
+
"epoch": 0.09961470299909068,
|
| 1627 |
+
"grad_norm": 0.19598425924777985,
|
| 1628 |
+
"learning_rate": 0.001,
|
| 1629 |
+
"loss": 3.221529235839844,
|
| 1630 |
+
"step": 11400
|
| 1631 |
+
},
|
| 1632 |
+
{
|
| 1633 |
+
"epoch": 0.10005160959119196,
|
| 1634 |
+
"grad_norm": 0.1995646357536316,
|
| 1635 |
+
"learning_rate": 0.001,
|
| 1636 |
+
"loss": 3.212632141113281,
|
| 1637 |
+
"step": 11450
|
| 1638 |
+
},
|
| 1639 |
+
{
|
| 1640 |
+
"epoch": 0.10048851618329324,
|
| 1641 |
+
"grad_norm": 0.1981515884399414,
|
| 1642 |
+
"learning_rate": 0.001,
|
| 1643 |
+
"loss": 3.2169342041015625,
|
| 1644 |
+
"step": 11500
|
| 1645 |
+
},
|
| 1646 |
+
{
|
| 1647 |
+
"epoch": 0.10092542277539451,
|
| 1648 |
+
"grad_norm": 0.19607605040073395,
|
| 1649 |
+
"learning_rate": 0.001,
|
| 1650 |
+
"loss": 3.212418212890625,
|
| 1651 |
+
"step": 11550
|
| 1652 |
+
},
|
| 1653 |
+
{
|
| 1654 |
+
"epoch": 0.10136232936749578,
|
| 1655 |
+
"grad_norm": 0.1733725219964981,
|
| 1656 |
+
"learning_rate": 0.001,
|
| 1657 |
+
"loss": 3.221158447265625,
|
| 1658 |
+
"step": 11600
|
| 1659 |
+
},
|
| 1660 |
+
{
|
| 1661 |
+
"epoch": 0.10179923595959706,
|
| 1662 |
+
"grad_norm": 0.18705779314041138,
|
| 1663 |
+
"learning_rate": 0.001,
|
| 1664 |
+
"loss": 3.2119464111328124,
|
| 1665 |
+
"step": 11650
|
| 1666 |
+
},
|
| 1667 |
+
{
|
| 1668 |
+
"epoch": 0.10223614255169834,
|
| 1669 |
+
"grad_norm": 0.19549337029457092,
|
| 1670 |
+
"learning_rate": 0.001,
|
| 1671 |
+
"loss": 3.2163186645507813,
|
| 1672 |
+
"step": 11700
|
| 1673 |
+
},
|
| 1674 |
+
{
|
| 1675 |
+
"epoch": 0.10267304914379961,
|
| 1676 |
+
"grad_norm": 0.21174098551273346,
|
| 1677 |
+
"learning_rate": 0.001,
|
| 1678 |
+
"loss": 3.22341552734375,
|
| 1679 |
+
"step": 11750
|
| 1680 |
+
},
|
| 1681 |
+
{
|
| 1682 |
+
"epoch": 0.10310995573590089,
|
| 1683 |
+
"grad_norm": 0.2125338762998581,
|
| 1684 |
+
"learning_rate": 0.001,
|
| 1685 |
+
"loss": 3.215631408691406,
|
| 1686 |
+
"step": 11800
|
| 1687 |
+
},
|
| 1688 |
+
{
|
| 1689 |
+
"epoch": 0.10354686232800216,
|
| 1690 |
+
"grad_norm": 0.1913818120956421,
|
| 1691 |
+
"learning_rate": 0.001,
|
| 1692 |
+
"loss": 3.215351867675781,
|
| 1693 |
+
"step": 11850
|
| 1694 |
+
},
|
| 1695 |
+
{
|
| 1696 |
+
"epoch": 0.10398376892010344,
|
| 1697 |
+
"grad_norm": 0.1795399934053421,
|
| 1698 |
+
"learning_rate": 0.001,
|
| 1699 |
+
"loss": 3.209778137207031,
|
| 1700 |
+
"step": 11900
|
| 1701 |
+
},
|
| 1702 |
+
{
|
| 1703 |
+
"epoch": 0.10442067551220471,
|
| 1704 |
+
"grad_norm": 0.18420544266700745,
|
| 1705 |
+
"learning_rate": 0.001,
|
| 1706 |
+
"loss": 3.214957275390625,
|
| 1707 |
+
"step": 11950
|
| 1708 |
+
},
|
| 1709 |
+
{
|
| 1710 |
+
"epoch": 0.10485758210430599,
|
| 1711 |
+
"grad_norm": 0.1968565434217453,
|
| 1712 |
+
"learning_rate": 0.001,
|
| 1713 |
+
"loss": 3.2145254516601565,
|
| 1714 |
+
"step": 12000
|
| 1715 |
+
},
|
| 1716 |
+
{
|
| 1717 |
+
"epoch": 0.10485758210430599,
|
| 1718 |
+
"eval_loss": 3.2212815284729004,
|
| 1719 |
+
"eval_runtime": 6.0309,
|
| 1720 |
+
"eval_samples_per_second": 161.835,
|
| 1721 |
+
"eval_steps_per_second": 10.115,
|
| 1722 |
+
"step": 12000
|
| 1723 |
+
},
|
| 1724 |
+
{
|
| 1725 |
+
"epoch": 0.10529448869640726,
|
| 1726 |
+
"grad_norm": 0.2099006026983261,
|
| 1727 |
+
"learning_rate": 0.001,
|
| 1728 |
+
"loss": 3.2097091674804688,
|
| 1729 |
+
"step": 12050
|
| 1730 |
+
},
|
| 1731 |
+
{
|
| 1732 |
+
"epoch": 0.10573139528850854,
|
| 1733 |
+
"grad_norm": 0.20223559439182281,
|
| 1734 |
+
"learning_rate": 0.001,
|
| 1735 |
+
"loss": 3.2140032958984377,
|
| 1736 |
+
"step": 12100
|
| 1737 |
+
},
|
| 1738 |
+
{
|
| 1739 |
+
"epoch": 0.10616830188060981,
|
| 1740 |
+
"grad_norm": 0.2003611922264099,
|
| 1741 |
+
"learning_rate": 0.001,
|
| 1742 |
+
"loss": 3.2128176879882813,
|
| 1743 |
+
"step": 12150
|
| 1744 |
+
},
|
| 1745 |
+
{
|
| 1746 |
+
"epoch": 0.10660520847271109,
|
| 1747 |
+
"grad_norm": 0.17242281138896942,
|
| 1748 |
+
"learning_rate": 0.001,
|
| 1749 |
+
"loss": 3.21211669921875,
|
| 1750 |
+
"step": 12200
|
| 1751 |
+
},
|
| 1752 |
+
{
|
| 1753 |
+
"epoch": 0.10704211506481236,
|
| 1754 |
+
"grad_norm": 0.18861070275306702,
|
| 1755 |
+
"learning_rate": 0.001,
|
| 1756 |
+
"loss": 3.2088800048828126,
|
| 1757 |
+
"step": 12250
|
| 1758 |
+
},
|
| 1759 |
+
{
|
| 1760 |
+
"epoch": 0.10747902165691364,
|
| 1761 |
+
"grad_norm": 0.1851765513420105,
|
| 1762 |
+
"learning_rate": 0.001,
|
| 1763 |
+
"loss": 3.2097900390625,
|
| 1764 |
+
"step": 12300
|
| 1765 |
+
},
|
| 1766 |
+
{
|
| 1767 |
+
"epoch": 0.10791592824901491,
|
| 1768 |
+
"grad_norm": 0.18498055636882782,
|
| 1769 |
+
"learning_rate": 0.001,
|
| 1770 |
+
"loss": 3.2144100952148436,
|
| 1771 |
+
"step": 12350
|
| 1772 |
+
},
|
| 1773 |
+
{
|
| 1774 |
+
"epoch": 0.10835283484111619,
|
| 1775 |
+
"grad_norm": 0.18698084354400635,
|
| 1776 |
+
"learning_rate": 0.001,
|
| 1777 |
+
"loss": 3.20849609375,
|
| 1778 |
+
"step": 12400
|
| 1779 |
+
},
|
| 1780 |
+
{
|
| 1781 |
+
"epoch": 0.10878974143321746,
|
| 1782 |
+
"grad_norm": 0.19474317133426666,
|
| 1783 |
+
"learning_rate": 0.001,
|
| 1784 |
+
"loss": 3.211873779296875,
|
| 1785 |
+
"step": 12450
|
| 1786 |
+
},
|
| 1787 |
+
{
|
| 1788 |
+
"epoch": 0.10922664802531874,
|
| 1789 |
+
"grad_norm": 0.17884086072444916,
|
| 1790 |
+
"learning_rate": 0.001,
|
| 1791 |
+
"loss": 3.2136822509765626,
|
| 1792 |
+
"step": 12500
|
| 1793 |
+
},
|
| 1794 |
+
{
|
| 1795 |
+
"epoch": 0.10966355461742001,
|
| 1796 |
+
"grad_norm": 0.19080394506454468,
|
| 1797 |
+
"learning_rate": 0.001,
|
| 1798 |
+
"loss": 3.2123370361328125,
|
| 1799 |
+
"step": 12550
|
| 1800 |
+
},
|
| 1801 |
+
{
|
| 1802 |
+
"epoch": 0.11010046120952129,
|
| 1803 |
+
"grad_norm": 0.1898597925901413,
|
| 1804 |
+
"learning_rate": 0.001,
|
| 1805 |
+
"loss": 3.2091488647460937,
|
| 1806 |
+
"step": 12600
|
| 1807 |
+
},
|
| 1808 |
+
{
|
| 1809 |
+
"epoch": 0.11053736780162256,
|
| 1810 |
+
"grad_norm": 0.18075662851333618,
|
| 1811 |
+
"learning_rate": 0.001,
|
| 1812 |
+
"loss": 3.2083639526367187,
|
| 1813 |
+
"step": 12650
|
| 1814 |
+
},
|
| 1815 |
+
{
|
| 1816 |
+
"epoch": 0.11097427439372384,
|
| 1817 |
+
"grad_norm": 0.1833634078502655,
|
| 1818 |
+
"learning_rate": 0.001,
|
| 1819 |
+
"loss": 3.2076034545898438,
|
| 1820 |
+
"step": 12700
|
| 1821 |
+
},
|
| 1822 |
+
{
|
| 1823 |
+
"epoch": 0.11141118098582511,
|
| 1824 |
+
"grad_norm": 0.19021724164485931,
|
| 1825 |
+
"learning_rate": 0.001,
|
| 1826 |
+
"loss": 3.203535461425781,
|
| 1827 |
+
"step": 12750
|
| 1828 |
+
},
|
| 1829 |
+
{
|
| 1830 |
+
"epoch": 0.11184808757792639,
|
| 1831 |
+
"grad_norm": 0.19512109458446503,
|
| 1832 |
+
"learning_rate": 0.001,
|
| 1833 |
+
"loss": 3.2149392700195314,
|
| 1834 |
+
"step": 12800
|
| 1835 |
+
},
|
| 1836 |
+
{
|
| 1837 |
+
"epoch": 0.11228499417002766,
|
| 1838 |
+
"grad_norm": 0.19543170928955078,
|
| 1839 |
+
"learning_rate": 0.001,
|
| 1840 |
+
"loss": 3.2061972045898437,
|
| 1841 |
+
"step": 12850
|
| 1842 |
+
},
|
| 1843 |
+
{
|
| 1844 |
+
"epoch": 0.11272190076212894,
|
| 1845 |
+
"grad_norm": 0.18879614770412445,
|
| 1846 |
+
"learning_rate": 0.001,
|
| 1847 |
+
"loss": 3.20185302734375,
|
| 1848 |
+
"step": 12900
|
| 1849 |
+
},
|
| 1850 |
+
{
|
| 1851 |
+
"epoch": 0.11315880735423021,
|
| 1852 |
+
"grad_norm": 0.17561078071594238,
|
| 1853 |
+
"learning_rate": 0.001,
|
| 1854 |
+
"loss": 3.19679443359375,
|
| 1855 |
+
"step": 12950
|
| 1856 |
+
},
|
| 1857 |
+
{
|
| 1858 |
+
"epoch": 0.11359571394633149,
|
| 1859 |
+
"grad_norm": 0.17411254346370697,
|
| 1860 |
+
"learning_rate": 0.001,
|
| 1861 |
+
"loss": 3.206636657714844,
|
| 1862 |
+
"step": 13000
|
| 1863 |
+
},
|
| 1864 |
+
{
|
| 1865 |
+
"epoch": 0.11403262053843276,
|
| 1866 |
+
"grad_norm": 0.19477669894695282,
|
| 1867 |
+
"learning_rate": 0.001,
|
| 1868 |
+
"loss": 3.2047265625,
|
| 1869 |
+
"step": 13050
|
| 1870 |
+
},
|
| 1871 |
+
{
|
| 1872 |
+
"epoch": 0.11446952713053403,
|
| 1873 |
+
"grad_norm": 0.2075432687997818,
|
| 1874 |
+
"learning_rate": 0.001,
|
| 1875 |
+
"loss": 3.2073565673828126,
|
| 1876 |
+
"step": 13100
|
| 1877 |
+
},
|
| 1878 |
+
{
|
| 1879 |
+
"epoch": 0.11490643372263531,
|
| 1880 |
+
"grad_norm": 0.19364356994628906,
|
| 1881 |
+
"learning_rate": 0.001,
|
| 1882 |
+
"loss": 3.2007147216796876,
|
| 1883 |
+
"step": 13150
|
| 1884 |
+
},
|
| 1885 |
+
{
|
| 1886 |
+
"epoch": 0.11534334031473659,
|
| 1887 |
+
"grad_norm": 0.19227240979671478,
|
| 1888 |
+
"learning_rate": 0.001,
|
| 1889 |
+
"loss": 3.2039303588867187,
|
| 1890 |
+
"step": 13200
|
| 1891 |
+
},
|
| 1892 |
+
{
|
| 1893 |
+
"epoch": 0.11578024690683786,
|
| 1894 |
+
"grad_norm": 0.1965664029121399,
|
| 1895 |
+
"learning_rate": 0.001,
|
| 1896 |
+
"loss": 3.204370422363281,
|
| 1897 |
+
"step": 13250
|
| 1898 |
+
},
|
| 1899 |
+
{
|
| 1900 |
+
"epoch": 0.11621715349893913,
|
| 1901 |
+
"grad_norm": 0.19625689089298248,
|
| 1902 |
+
"learning_rate": 0.001,
|
| 1903 |
+
"loss": 3.2018893432617186,
|
| 1904 |
+
"step": 13300
|
| 1905 |
+
},
|
| 1906 |
+
{
|
| 1907 |
+
"epoch": 0.11665406009104042,
|
| 1908 |
+
"grad_norm": 0.1872740089893341,
|
| 1909 |
+
"learning_rate": 0.001,
|
| 1910 |
+
"loss": 3.1984957885742187,
|
| 1911 |
+
"step": 13350
|
| 1912 |
+
},
|
| 1913 |
+
{
|
| 1914 |
+
"epoch": 0.11709096668314169,
|
| 1915 |
+
"grad_norm": 0.17404606938362122,
|
| 1916 |
+
"learning_rate": 0.001,
|
| 1917 |
+
"loss": 3.2010247802734373,
|
| 1918 |
+
"step": 13400
|
| 1919 |
+
},
|
| 1920 |
+
{
|
| 1921 |
+
"epoch": 0.11752787327524296,
|
| 1922 |
+
"grad_norm": 0.20951177179813385,
|
| 1923 |
+
"learning_rate": 0.001,
|
| 1924 |
+
"loss": 3.2027737426757814,
|
| 1925 |
+
"step": 13450
|
| 1926 |
+
},
|
| 1927 |
+
{
|
| 1928 |
+
"epoch": 0.11796477986734423,
|
| 1929 |
+
"grad_norm": 0.20374071598052979,
|
| 1930 |
+
"learning_rate": 0.001,
|
| 1931 |
+
"loss": 3.1948858642578126,
|
| 1932 |
+
"step": 13500
|
| 1933 |
+
},
|
| 1934 |
+
{
|
| 1935 |
+
"epoch": 0.11840168645944552,
|
| 1936 |
+
"grad_norm": 0.21140068769454956,
|
| 1937 |
+
"learning_rate": 0.001,
|
| 1938 |
+
"loss": 3.206798095703125,
|
| 1939 |
+
"step": 13550
|
| 1940 |
+
},
|
| 1941 |
+
{
|
| 1942 |
+
"epoch": 0.11883859305154679,
|
| 1943 |
+
"grad_norm": 0.17180323600769043,
|
| 1944 |
+
"learning_rate": 0.001,
|
| 1945 |
+
"loss": 3.197232666015625,
|
| 1946 |
+
"step": 13600
|
| 1947 |
+
},
|
| 1948 |
+
{
|
| 1949 |
+
"epoch": 0.11927549964364806,
|
| 1950 |
+
"grad_norm": 0.1982830911874771,
|
| 1951 |
+
"learning_rate": 0.001,
|
| 1952 |
+
"loss": 3.196986083984375,
|
| 1953 |
+
"step": 13650
|
| 1954 |
+
},
|
| 1955 |
+
{
|
| 1956 |
+
"epoch": 0.11971240623574933,
|
| 1957 |
+
"grad_norm": 0.19306933879852295,
|
| 1958 |
+
"learning_rate": 0.001,
|
| 1959 |
+
"loss": 3.1933523559570314,
|
| 1960 |
+
"step": 13700
|
| 1961 |
+
},
|
| 1962 |
+
{
|
| 1963 |
+
"epoch": 0.12014931282785062,
|
| 1964 |
+
"grad_norm": 0.21267230808734894,
|
| 1965 |
+
"learning_rate": 0.001,
|
| 1966 |
+
"loss": 3.2039981079101563,
|
| 1967 |
+
"step": 13750
|
| 1968 |
+
},
|
| 1969 |
+
{
|
| 1970 |
+
"epoch": 0.12058621941995189,
|
| 1971 |
+
"grad_norm": 0.2193242907524109,
|
| 1972 |
+
"learning_rate": 0.001,
|
| 1973 |
+
"loss": 3.1955245971679687,
|
| 1974 |
+
"step": 13800
|
| 1975 |
+
},
|
| 1976 |
+
{
|
| 1977 |
+
"epoch": 0.12102312601205316,
|
| 1978 |
+
"grad_norm": 0.23339790105819702,
|
| 1979 |
+
"learning_rate": 0.001,
|
| 1980 |
+
"loss": 3.1981427001953127,
|
| 1981 |
+
"step": 13850
|
| 1982 |
+
},
|
| 1983 |
+
{
|
| 1984 |
+
"epoch": 0.12146003260415443,
|
| 1985 |
+
"grad_norm": 0.22573474049568176,
|
| 1986 |
+
"learning_rate": 0.001,
|
| 1987 |
+
"loss": 3.196231689453125,
|
| 1988 |
+
"step": 13900
|
| 1989 |
+
},
|
| 1990 |
+
{
|
| 1991 |
+
"epoch": 0.12189693919625572,
|
| 1992 |
+
"grad_norm": 0.19238780438899994,
|
| 1993 |
+
"learning_rate": 0.001,
|
| 1994 |
+
"loss": 3.1946224975585937,
|
| 1995 |
+
"step": 13950
|
| 1996 |
+
},
|
| 1997 |
+
{
|
| 1998 |
+
"epoch": 0.12233384578835699,
|
| 1999 |
+
"grad_norm": 0.202445387840271,
|
| 2000 |
+
"learning_rate": 0.001,
|
| 2001 |
+
"loss": 3.194338073730469,
|
| 2002 |
+
"step": 14000
|
| 2003 |
+
},
|
| 2004 |
+
{
|
| 2005 |
+
"epoch": 0.12277075238045826,
|
| 2006 |
+
"grad_norm": 0.18492762744426727,
|
| 2007 |
+
"learning_rate": 0.001,
|
| 2008 |
+
"loss": 3.2063372802734373,
|
| 2009 |
+
"step": 14050
|
| 2010 |
+
},
|
| 2011 |
+
{
|
| 2012 |
+
"epoch": 0.12320765897255953,
|
| 2013 |
+
"grad_norm": 0.20408518612384796,
|
| 2014 |
+
"learning_rate": 0.001,
|
| 2015 |
+
"loss": 3.1926824951171877,
|
| 2016 |
+
"step": 14100
|
| 2017 |
+
},
|
| 2018 |
+
{
|
| 2019 |
+
"epoch": 0.12364456556466082,
|
| 2020 |
+
"grad_norm": 0.19603824615478516,
|
| 2021 |
+
"learning_rate": 0.001,
|
| 2022 |
+
"loss": 3.1970639038085937,
|
| 2023 |
+
"step": 14150
|
| 2024 |
+
},
|
| 2025 |
+
{
|
| 2026 |
+
"epoch": 0.12408147215676209,
|
| 2027 |
+
"grad_norm": 0.18211990594863892,
|
| 2028 |
+
"learning_rate": 0.001,
|
| 2029 |
+
"loss": 3.193045654296875,
|
| 2030 |
+
"step": 14200
|
| 2031 |
+
},
|
| 2032 |
+
{
|
| 2033 |
+
"epoch": 0.12451837874886336,
|
| 2034 |
+
"grad_norm": 0.21678605675697327,
|
| 2035 |
+
"learning_rate": 0.001,
|
| 2036 |
+
"loss": 3.1917852783203124,
|
| 2037 |
+
"step": 14250
|
| 2038 |
+
},
|
| 2039 |
+
{
|
| 2040 |
+
"epoch": 0.12495528534096463,
|
| 2041 |
+
"grad_norm": 0.19074733555316925,
|
| 2042 |
+
"learning_rate": 0.001,
|
| 2043 |
+
"loss": 3.1935299682617186,
|
| 2044 |
+
"step": 14300
|
| 2045 |
+
},
|
| 2046 |
+
{
|
| 2047 |
+
"epoch": 0.12539219193306592,
|
| 2048 |
+
"grad_norm": 0.17810741066932678,
|
| 2049 |
+
"learning_rate": 0.001,
|
| 2050 |
+
"loss": 3.1906802368164064,
|
| 2051 |
+
"step": 14350
|
| 2052 |
+
},
|
| 2053 |
+
{
|
| 2054 |
+
"epoch": 0.1258290985251672,
|
| 2055 |
+
"grad_norm": 0.19822879135608673,
|
| 2056 |
+
"learning_rate": 0.001,
|
| 2057 |
+
"loss": 3.1980941772460936,
|
| 2058 |
+
"step": 14400
|
| 2059 |
+
},
|
| 2060 |
+
{
|
| 2061 |
+
"epoch": 0.12626600511726846,
|
| 2062 |
+
"grad_norm": 0.20151387155056,
|
| 2063 |
+
"learning_rate": 0.001,
|
| 2064 |
+
"loss": 3.1888507080078123,
|
| 2065 |
+
"step": 14450
|
| 2066 |
+
},
|
| 2067 |
+
{
|
| 2068 |
+
"epoch": 0.12670291170936973,
|
| 2069 |
+
"grad_norm": 0.1979341208934784,
|
| 2070 |
+
"learning_rate": 0.001,
|
| 2071 |
+
"loss": 3.1934823608398437,
|
| 2072 |
+
"step": 14500
|
| 2073 |
+
},
|
| 2074 |
+
{
|
| 2075 |
+
"epoch": 0.127139818301471,
|
| 2076 |
+
"grad_norm": 0.1844247728586197,
|
| 2077 |
+
"learning_rate": 0.001,
|
| 2078 |
+
"loss": 3.1945904541015624,
|
| 2079 |
+
"step": 14550
|
| 2080 |
+
},
|
| 2081 |
+
{
|
| 2082 |
+
"epoch": 0.12757672489357227,
|
| 2083 |
+
"grad_norm": 0.1843341737985611,
|
| 2084 |
+
"learning_rate": 0.001,
|
| 2085 |
+
"loss": 3.1908944702148436,
|
| 2086 |
+
"step": 14600
|
| 2087 |
+
},
|
| 2088 |
+
{
|
| 2089 |
+
"epoch": 0.12801363148567357,
|
| 2090 |
+
"grad_norm": 0.19181329011917114,
|
| 2091 |
+
"learning_rate": 0.001,
|
| 2092 |
+
"loss": 3.186395263671875,
|
| 2093 |
+
"step": 14650
|
| 2094 |
+
},
|
| 2095 |
+
{
|
| 2096 |
+
"epoch": 0.12845053807777485,
|
| 2097 |
+
"grad_norm": 0.1925257444381714,
|
| 2098 |
+
"learning_rate": 0.001,
|
| 2099 |
+
"loss": 3.187642822265625,
|
| 2100 |
+
"step": 14700
|
| 2101 |
+
},
|
| 2102 |
+
{
|
| 2103 |
+
"epoch": 0.12888744466987612,
|
| 2104 |
+
"grad_norm": 0.17801596224308014,
|
| 2105 |
+
"learning_rate": 0.001,
|
| 2106 |
+
"loss": 3.18369140625,
|
| 2107 |
+
"step": 14750
|
| 2108 |
+
},
|
| 2109 |
+
{
|
| 2110 |
+
"epoch": 0.1293243512619774,
|
| 2111 |
+
"grad_norm": 0.20504137873649597,
|
| 2112 |
+
"learning_rate": 0.001,
|
| 2113 |
+
"loss": 3.1881332397460938,
|
| 2114 |
+
"step": 14800
|
| 2115 |
+
},
|
| 2116 |
+
{
|
| 2117 |
+
"epoch": 0.12976125785407866,
|
| 2118 |
+
"grad_norm": 0.18415763974189758,
|
| 2119 |
+
"learning_rate": 0.001,
|
| 2120 |
+
"loss": 3.1859930419921874,
|
| 2121 |
+
"step": 14850
|
| 2122 |
+
},
|
| 2123 |
+
{
|
| 2124 |
+
"epoch": 0.13019816444617993,
|
| 2125 |
+
"grad_norm": 0.2078758180141449,
|
| 2126 |
+
"learning_rate": 0.001,
|
| 2127 |
+
"loss": 3.19252197265625,
|
| 2128 |
+
"step": 14900
|
| 2129 |
+
},
|
| 2130 |
+
{
|
| 2131 |
+
"epoch": 0.1306350710382812,
|
| 2132 |
+
"grad_norm": 0.1912207156419754,
|
| 2133 |
+
"learning_rate": 0.001,
|
| 2134 |
+
"loss": 3.1930859375,
|
| 2135 |
+
"step": 14950
|
| 2136 |
+
},
|
| 2137 |
+
{
|
| 2138 |
+
"epoch": 0.13107197763038247,
|
| 2139 |
+
"grad_norm": 0.18659965693950653,
|
| 2140 |
+
"learning_rate": 0.001,
|
| 2141 |
+
"loss": 3.1894403076171876,
|
| 2142 |
+
"step": 15000
|
| 2143 |
+
},
|
| 2144 |
+
{
|
| 2145 |
+
"epoch": 0.13107197763038247,
|
| 2146 |
+
"eval_loss": 3.1984639167785645,
|
| 2147 |
+
"eval_runtime": 9.5838,
|
| 2148 |
+
"eval_samples_per_second": 101.838,
|
| 2149 |
+
"eval_steps_per_second": 6.365,
|
| 2150 |
+
"step": 15000
|
| 2151 |
+
},
|
| 2152 |
+
{
|
| 2153 |
+
"epoch": 0.13150888422248375,
|
| 2154 |
+
"grad_norm": 0.18150147795677185,
|
| 2155 |
+
"learning_rate": 0.001,
|
| 2156 |
+
"loss": 3.1855987548828124,
|
| 2157 |
+
"step": 15050
|
| 2158 |
+
},
|
| 2159 |
+
{
|
| 2160 |
+
"epoch": 0.13194579081458505,
|
| 2161 |
+
"grad_norm": 0.1850251406431198,
|
| 2162 |
+
"learning_rate": 0.001,
|
| 2163 |
+
"loss": 3.18405517578125,
|
| 2164 |
+
"step": 15100
|
| 2165 |
+
},
|
| 2166 |
+
{
|
| 2167 |
+
"epoch": 0.13238269740668632,
|
| 2168 |
+
"grad_norm": 0.18930986523628235,
|
| 2169 |
+
"learning_rate": 0.001,
|
| 2170 |
+
"loss": 3.181584777832031,
|
| 2171 |
+
"step": 15150
|
| 2172 |
+
},
|
| 2173 |
+
{
|
| 2174 |
+
"epoch": 0.1328196039987876,
|
| 2175 |
+
"grad_norm": 0.19745807349681854,
|
| 2176 |
+
"learning_rate": 0.001,
|
| 2177 |
+
"loss": 3.1843905639648438,
|
| 2178 |
+
"step": 15200
|
| 2179 |
+
},
|
| 2180 |
+
{
|
| 2181 |
+
"epoch": 0.13325651059088886,
|
| 2182 |
+
"grad_norm": 0.2121584564447403,
|
| 2183 |
+
"learning_rate": 0.001,
|
| 2184 |
+
"loss": 3.194527893066406,
|
| 2185 |
+
"step": 15250
|
| 2186 |
+
},
|
| 2187 |
+
{
|
| 2188 |
+
"epoch": 0.13369341718299013,
|
| 2189 |
+
"grad_norm": 0.1978662610054016,
|
| 2190 |
+
"learning_rate": 0.001,
|
| 2191 |
+
"loss": 3.185810241699219,
|
| 2192 |
+
"step": 15300
|
| 2193 |
+
},
|
| 2194 |
+
{
|
| 2195 |
+
"epoch": 0.1341303237750914,
|
| 2196 |
+
"grad_norm": 0.19299598038196564,
|
| 2197 |
+
"learning_rate": 0.001,
|
| 2198 |
+
"loss": 3.178111877441406,
|
| 2199 |
+
"step": 15350
|
| 2200 |
+
},
|
| 2201 |
+
{
|
| 2202 |
+
"epoch": 0.13456723036719268,
|
| 2203 |
+
"grad_norm": 0.1843535304069519,
|
| 2204 |
+
"learning_rate": 0.001,
|
| 2205 |
+
"loss": 3.185999450683594,
|
| 2206 |
+
"step": 15400
|
| 2207 |
+
},
|
| 2208 |
+
{
|
| 2209 |
+
"epoch": 0.13500413695929395,
|
| 2210 |
+
"grad_norm": 0.17675578594207764,
|
| 2211 |
+
"learning_rate": 0.001,
|
| 2212 |
+
"loss": 3.1846661376953125,
|
| 2213 |
+
"step": 15450
|
| 2214 |
+
},
|
| 2215 |
+
{
|
| 2216 |
+
"epoch": 0.13544104355139525,
|
| 2217 |
+
"grad_norm": 0.19691866636276245,
|
| 2218 |
+
"learning_rate": 0.001,
|
| 2219 |
+
"loss": 3.184285888671875,
|
| 2220 |
+
"step": 15500
|
| 2221 |
+
},
|
| 2222 |
+
{
|
| 2223 |
+
"epoch": 0.13587795014349652,
|
| 2224 |
+
"grad_norm": 0.23492923378944397,
|
| 2225 |
+
"learning_rate": 0.001,
|
| 2226 |
+
"loss": 3.1907785034179685,
|
| 2227 |
+
"step": 15550
|
| 2228 |
+
},
|
| 2229 |
+
{
|
| 2230 |
+
"epoch": 0.1363148567355978,
|
| 2231 |
+
"grad_norm": 0.1797175109386444,
|
| 2232 |
+
"learning_rate": 0.001,
|
| 2233 |
+
"loss": 3.184339599609375,
|
| 2234 |
+
"step": 15600
|
| 2235 |
+
},
|
| 2236 |
+
{
|
| 2237 |
+
"epoch": 0.13675176332769906,
|
| 2238 |
+
"grad_norm": 0.18342596292495728,
|
| 2239 |
+
"learning_rate": 0.001,
|
| 2240 |
+
"loss": 3.1823294067382815,
|
| 2241 |
+
"step": 15650
|
| 2242 |
+
},
|
| 2243 |
+
{
|
| 2244 |
+
"epoch": 0.13718866991980033,
|
| 2245 |
+
"grad_norm": 0.20209701359272003,
|
| 2246 |
+
"learning_rate": 0.001,
|
| 2247 |
+
"loss": 3.187270202636719,
|
| 2248 |
+
"step": 15700
|
| 2249 |
+
},
|
| 2250 |
+
{
|
| 2251 |
+
"epoch": 0.1376255765119016,
|
| 2252 |
+
"grad_norm": 0.17159967124462128,
|
| 2253 |
+
"learning_rate": 0.001,
|
| 2254 |
+
"loss": 3.187642822265625,
|
| 2255 |
+
"step": 15750
|
| 2256 |
+
},
|
| 2257 |
+
{
|
| 2258 |
+
"epoch": 0.13806248310400288,
|
| 2259 |
+
"grad_norm": 0.18459446728229523,
|
| 2260 |
+
"learning_rate": 0.001,
|
| 2261 |
+
"loss": 3.196175537109375,
|
| 2262 |
+
"step": 15800
|
| 2263 |
+
},
|
| 2264 |
+
{
|
| 2265 |
+
"epoch": 0.13849938969610415,
|
| 2266 |
+
"grad_norm": 0.21580736339092255,
|
| 2267 |
+
"learning_rate": 0.001,
|
| 2268 |
+
"loss": 3.179403076171875,
|
| 2269 |
+
"step": 15850
|
| 2270 |
+
},
|
| 2271 |
+
{
|
| 2272 |
+
"epoch": 0.13893629628820545,
|
| 2273 |
+
"grad_norm": 0.20760612189769745,
|
| 2274 |
+
"learning_rate": 0.001,
|
| 2275 |
+
"loss": 3.1841162109375,
|
| 2276 |
+
"step": 15900
|
| 2277 |
+
},
|
| 2278 |
+
{
|
| 2279 |
+
"epoch": 0.13937320288030672,
|
| 2280 |
+
"grad_norm": 0.19794167578220367,
|
| 2281 |
+
"learning_rate": 0.001,
|
| 2282 |
+
"loss": 3.1841552734375,
|
| 2283 |
+
"step": 15950
|
| 2284 |
+
},
|
| 2285 |
+
{
|
| 2286 |
+
"epoch": 0.139810109472408,
|
| 2287 |
+
"grad_norm": 0.19085007905960083,
|
| 2288 |
+
"learning_rate": 0.001,
|
| 2289 |
+
"loss": 3.179669189453125,
|
| 2290 |
+
"step": 16000
|
| 2291 |
+
},
|
| 2292 |
+
{
|
| 2293 |
+
"epoch": 0.14024701606450926,
|
| 2294 |
+
"grad_norm": 0.18777291476726532,
|
| 2295 |
+
"learning_rate": 0.001,
|
| 2296 |
+
"loss": 3.1862728881835936,
|
| 2297 |
+
"step": 16050
|
| 2298 |
+
},
|
| 2299 |
+
{
|
| 2300 |
+
"epoch": 0.14068392265661053,
|
| 2301 |
+
"grad_norm": 0.1883874386548996,
|
| 2302 |
+
"learning_rate": 0.001,
|
| 2303 |
+
"loss": 3.1826882934570313,
|
| 2304 |
+
"step": 16100
|
| 2305 |
+
},
|
| 2306 |
+
{
|
| 2307 |
+
"epoch": 0.1411208292487118,
|
| 2308 |
+
"grad_norm": 0.1788993775844574,
|
| 2309 |
+
"learning_rate": 0.001,
|
| 2310 |
+
"loss": 3.1862246704101564,
|
| 2311 |
+
"step": 16150
|
| 2312 |
+
},
|
| 2313 |
+
{
|
| 2314 |
+
"epoch": 0.14155773584081308,
|
| 2315 |
+
"grad_norm": 0.19650617241859436,
|
| 2316 |
+
"learning_rate": 0.001,
|
| 2317 |
+
"loss": 3.1779071044921876,
|
| 2318 |
+
"step": 16200
|
| 2319 |
+
},
|
| 2320 |
+
{
|
| 2321 |
+
"epoch": 0.14199464243291435,
|
| 2322 |
+
"grad_norm": 0.17883451282978058,
|
| 2323 |
+
"learning_rate": 0.001,
|
| 2324 |
+
"loss": 3.1806765747070314,
|
| 2325 |
+
"step": 16250
|
| 2326 |
+
},
|
| 2327 |
+
{
|
| 2328 |
+
"epoch": 0.14243154902501565,
|
| 2329 |
+
"grad_norm": 0.22816094756126404,
|
| 2330 |
+
"learning_rate": 0.001,
|
| 2331 |
+
"loss": 3.179928894042969,
|
| 2332 |
+
"step": 16300
|
| 2333 |
+
},
|
| 2334 |
+
{
|
| 2335 |
+
"epoch": 0.14286845561711692,
|
| 2336 |
+
"grad_norm": 0.18868671357631683,
|
| 2337 |
+
"learning_rate": 0.001,
|
| 2338 |
+
"loss": 3.180938720703125,
|
| 2339 |
+
"step": 16350
|
| 2340 |
+
},
|
| 2341 |
+
{
|
| 2342 |
+
"epoch": 0.1433053622092182,
|
| 2343 |
+
"grad_norm": 0.18312901258468628,
|
| 2344 |
+
"learning_rate": 0.001,
|
| 2345 |
+
"loss": 3.1809161376953123,
|
| 2346 |
+
"step": 16400
|
| 2347 |
+
},
|
| 2348 |
+
{
|
| 2349 |
+
"epoch": 0.14374226880131946,
|
| 2350 |
+
"grad_norm": 0.18817898631095886,
|
| 2351 |
+
"learning_rate": 0.001,
|
| 2352 |
+
"loss": 3.1803634643554686,
|
| 2353 |
+
"step": 16450
|
| 2354 |
+
},
|
| 2355 |
+
{
|
| 2356 |
+
"epoch": 0.14417917539342073,
|
| 2357 |
+
"grad_norm": 0.1878128945827484,
|
| 2358 |
+
"learning_rate": 0.001,
|
| 2359 |
+
"loss": 3.184964294433594,
|
| 2360 |
+
"step": 16500
|
| 2361 |
+
},
|
| 2362 |
+
{
|
| 2363 |
+
"epoch": 0.144616081985522,
|
| 2364 |
+
"grad_norm": 0.19661137461662292,
|
| 2365 |
+
"learning_rate": 0.001,
|
| 2366 |
+
"loss": 3.1799737548828126,
|
| 2367 |
+
"step": 16550
|
| 2368 |
+
},
|
| 2369 |
+
{
|
| 2370 |
+
"epoch": 0.14505298857762328,
|
| 2371 |
+
"grad_norm": 0.19748558104038239,
|
| 2372 |
+
"learning_rate": 0.001,
|
| 2373 |
+
"loss": 3.174441223144531,
|
| 2374 |
+
"step": 16600
|
| 2375 |
+
},
|
| 2376 |
+
{
|
| 2377 |
+
"epoch": 0.14548989516972455,
|
| 2378 |
+
"grad_norm": 0.18286362290382385,
|
| 2379 |
+
"learning_rate": 0.001,
|
| 2380 |
+
"loss": 3.1855096435546875,
|
| 2381 |
+
"step": 16650
|
| 2382 |
+
},
|
| 2383 |
+
{
|
| 2384 |
+
"epoch": 0.14592680176182582,
|
| 2385 |
+
"grad_norm": 0.20769278705120087,
|
| 2386 |
+
"learning_rate": 0.001,
|
| 2387 |
+
"loss": 3.1789508056640625,
|
| 2388 |
+
"step": 16700
|
| 2389 |
+
},
|
| 2390 |
+
{
|
| 2391 |
+
"epoch": 0.14636370835392712,
|
| 2392 |
+
"grad_norm": 0.19547678530216217,
|
| 2393 |
+
"learning_rate": 0.001,
|
| 2394 |
+
"loss": 3.180931396484375,
|
| 2395 |
+
"step": 16750
|
| 2396 |
+
},
|
| 2397 |
+
{
|
| 2398 |
+
"epoch": 0.1468006149460284,
|
| 2399 |
+
"grad_norm": 0.18609896302223206,
|
| 2400 |
+
"learning_rate": 0.001,
|
| 2401 |
+
"loss": 3.1838433837890623,
|
| 2402 |
+
"step": 16800
|
| 2403 |
+
},
|
| 2404 |
+
{
|
| 2405 |
+
"epoch": 0.14723752153812966,
|
| 2406 |
+
"grad_norm": 0.19330042600631714,
|
| 2407 |
+
"learning_rate": 0.001,
|
| 2408 |
+
"loss": 3.1765182495117186,
|
| 2409 |
+
"step": 16850
|
| 2410 |
+
},
|
| 2411 |
+
{
|
| 2412 |
+
"epoch": 0.14767442813023093,
|
| 2413 |
+
"grad_norm": 0.18436136841773987,
|
| 2414 |
+
"learning_rate": 0.001,
|
| 2415 |
+
"loss": 3.17610595703125,
|
| 2416 |
+
"step": 16900
|
| 2417 |
+
},
|
| 2418 |
+
{
|
| 2419 |
+
"epoch": 0.1481113347223322,
|
| 2420 |
+
"grad_norm": 0.18651925027370453,
|
| 2421 |
+
"learning_rate": 0.001,
|
| 2422 |
+
"loss": 3.1826718139648436,
|
| 2423 |
+
"step": 16950
|
| 2424 |
+
},
|
| 2425 |
+
{
|
| 2426 |
+
"epoch": 0.14854824131443348,
|
| 2427 |
+
"grad_norm": 0.1850690394639969,
|
| 2428 |
+
"learning_rate": 0.001,
|
| 2429 |
+
"loss": 3.172263488769531,
|
| 2430 |
+
"step": 17000
|
| 2431 |
+
},
|
| 2432 |
+
{
|
| 2433 |
+
"epoch": 0.14898514790653475,
|
| 2434 |
+
"grad_norm": 0.19027315080165863,
|
| 2435 |
+
"learning_rate": 0.001,
|
| 2436 |
+
"loss": 3.174286193847656,
|
| 2437 |
+
"step": 17050
|
| 2438 |
+
},
|
| 2439 |
+
{
|
| 2440 |
+
"epoch": 0.14942205449863602,
|
| 2441 |
+
"grad_norm": 0.1988609880208969,
|
| 2442 |
+
"learning_rate": 0.001,
|
| 2443 |
+
"loss": 3.1703976440429686,
|
| 2444 |
+
"step": 17100
|
| 2445 |
+
},
|
| 2446 |
+
{
|
| 2447 |
+
"epoch": 0.14985896109073732,
|
| 2448 |
+
"grad_norm": 0.1764635443687439,
|
| 2449 |
+
"learning_rate": 0.001,
|
| 2450 |
+
"loss": 3.1757376098632815,
|
| 2451 |
+
"step": 17150
|
| 2452 |
+
},
|
| 2453 |
+
{
|
| 2454 |
+
"epoch": 0.1502958676828386,
|
| 2455 |
+
"grad_norm": 0.17891408503055573,
|
| 2456 |
+
"learning_rate": 0.001,
|
| 2457 |
+
"loss": 3.179452209472656,
|
| 2458 |
+
"step": 17200
|
| 2459 |
+
},
|
| 2460 |
+
{
|
| 2461 |
+
"epoch": 0.15073277427493986,
|
| 2462 |
+
"grad_norm": 0.18927693367004395,
|
| 2463 |
+
"learning_rate": 0.001,
|
| 2464 |
+
"loss": 3.1811477661132814,
|
| 2465 |
+
"step": 17250
|
| 2466 |
+
},
|
| 2467 |
+
{
|
| 2468 |
+
"epoch": 0.15116968086704113,
|
| 2469 |
+
"grad_norm": 0.19046278297901154,
|
| 2470 |
+
"learning_rate": 0.001,
|
| 2471 |
+
"loss": 3.1781185913085936,
|
| 2472 |
+
"step": 17300
|
| 2473 |
+
},
|
| 2474 |
+
{
|
| 2475 |
+
"epoch": 0.1516065874591424,
|
| 2476 |
+
"grad_norm": 0.2253022938966751,
|
| 2477 |
+
"learning_rate": 0.001,
|
| 2478 |
+
"loss": 3.1700704956054686,
|
| 2479 |
+
"step": 17350
|
| 2480 |
+
},
|
| 2481 |
+
{
|
| 2482 |
+
"epoch": 0.15204349405124368,
|
| 2483 |
+
"grad_norm": 0.18004675209522247,
|
| 2484 |
+
"learning_rate": 0.001,
|
| 2485 |
+
"loss": 3.174186096191406,
|
| 2486 |
+
"step": 17400
|
| 2487 |
+
},
|
| 2488 |
+
{
|
| 2489 |
+
"epoch": 0.15248040064334495,
|
| 2490 |
+
"grad_norm": 0.18288516998291016,
|
| 2491 |
+
"learning_rate": 0.001,
|
| 2492 |
+
"loss": 3.1777203369140623,
|
| 2493 |
+
"step": 17450
|
| 2494 |
+
},
|
| 2495 |
+
{
|
| 2496 |
+
"epoch": 0.15291730723544622,
|
| 2497 |
+
"grad_norm": 0.18469488620758057,
|
| 2498 |
+
"learning_rate": 0.001,
|
| 2499 |
+
"loss": 3.1725430297851562,
|
| 2500 |
+
"step": 17500
|
| 2501 |
+
},
|
| 2502 |
+
{
|
| 2503 |
+
"epoch": 0.15335421382754752,
|
| 2504 |
+
"grad_norm": 0.19650670886039734,
|
| 2505 |
+
"learning_rate": 0.001,
|
| 2506 |
+
"loss": 3.180466613769531,
|
| 2507 |
+
"step": 17550
|
| 2508 |
+
},
|
| 2509 |
+
{
|
| 2510 |
+
"epoch": 0.1537911204196488,
|
| 2511 |
+
"grad_norm": 0.18393175303936005,
|
| 2512 |
+
"learning_rate": 0.001,
|
| 2513 |
+
"loss": 3.1719656372070313,
|
| 2514 |
+
"step": 17600
|
| 2515 |
+
},
|
| 2516 |
+
{
|
| 2517 |
+
"epoch": 0.15422802701175006,
|
| 2518 |
+
"grad_norm": 0.22304636240005493,
|
| 2519 |
+
"learning_rate": 0.001,
|
| 2520 |
+
"loss": 3.176797180175781,
|
| 2521 |
+
"step": 17650
|
| 2522 |
+
},
|
| 2523 |
+
{
|
| 2524 |
+
"epoch": 0.15466493360385133,
|
| 2525 |
+
"grad_norm": 0.19842109084129333,
|
| 2526 |
+
"learning_rate": 0.001,
|
| 2527 |
+
"loss": 3.1707284545898435,
|
| 2528 |
+
"step": 17700
|
| 2529 |
+
},
|
| 2530 |
+
{
|
| 2531 |
+
"epoch": 0.1551018401959526,
|
| 2532 |
+
"grad_norm": 0.17771974205970764,
|
| 2533 |
+
"learning_rate": 0.001,
|
| 2534 |
+
"loss": 3.1734857177734375,
|
| 2535 |
+
"step": 17750
|
| 2536 |
+
},
|
| 2537 |
+
{
|
| 2538 |
+
"epoch": 0.15553874678805388,
|
| 2539 |
+
"grad_norm": 0.17204603552818298,
|
| 2540 |
+
"learning_rate": 0.001,
|
| 2541 |
+
"loss": 3.170624084472656,
|
| 2542 |
+
"step": 17800
|
| 2543 |
+
},
|
| 2544 |
+
{
|
| 2545 |
+
"epoch": 0.15597565338015515,
|
| 2546 |
+
"grad_norm": 0.22022388875484467,
|
| 2547 |
+
"learning_rate": 0.001,
|
| 2548 |
+
"loss": 3.180273132324219,
|
| 2549 |
+
"step": 17850
|
| 2550 |
+
},
|
| 2551 |
+
{
|
| 2552 |
+
"epoch": 0.15641255997225642,
|
| 2553 |
+
"grad_norm": 0.183755025267601,
|
| 2554 |
+
"learning_rate": 0.001,
|
| 2555 |
+
"loss": 3.177940673828125,
|
| 2556 |
+
"step": 17900
|
| 2557 |
+
},
|
| 2558 |
+
{
|
| 2559 |
+
"epoch": 0.15684946656435772,
|
| 2560 |
+
"grad_norm": 0.20767413079738617,
|
| 2561 |
+
"learning_rate": 0.001,
|
| 2562 |
+
"loss": 3.17085205078125,
|
| 2563 |
+
"step": 17950
|
| 2564 |
+
},
|
| 2565 |
+
{
|
| 2566 |
+
"epoch": 0.157286373156459,
|
| 2567 |
+
"grad_norm": 0.1834259182214737,
|
| 2568 |
+
"learning_rate": 0.001,
|
| 2569 |
+
"loss": 3.168044128417969,
|
| 2570 |
+
"step": 18000
|
| 2571 |
+
},
|
| 2572 |
+
{
|
| 2573 |
+
"epoch": 0.157286373156459,
|
| 2574 |
+
"eval_loss": 3.180673837661743,
|
| 2575 |
+
"eval_runtime": 6.2659,
|
| 2576 |
+
"eval_samples_per_second": 155.763,
|
| 2577 |
+
"eval_steps_per_second": 9.735,
|
| 2578 |
+
"step": 18000
|
| 2579 |
+
}
|
| 2580 |
+
],
|
| 2581 |
+
"logging_steps": 50,
|
| 2582 |
+
"max_steps": 114440,
|
| 2583 |
+
"num_input_tokens_seen": 0,
|
| 2584 |
+
"num_train_epochs": 1,
|
| 2585 |
+
"save_steps": 3000,
|
| 2586 |
+
"stateful_callbacks": {
|
| 2587 |
+
"TrainerControl": {
|
| 2588 |
+
"args": {
|
| 2589 |
+
"should_epoch_stop": false,
|
| 2590 |
+
"should_evaluate": false,
|
| 2591 |
+
"should_log": false,
|
| 2592 |
+
"should_save": true,
|
| 2593 |
+
"should_training_stop": false
|
| 2594 |
+
},
|
| 2595 |
+
"attributes": {}
|
| 2596 |
+
}
|
| 2597 |
+
},
|
| 2598 |
+
"total_flos": 2.138037672738816e+18,
|
| 2599 |
+
"train_batch_size": 16,
|
| 2600 |
+
"trial_name": null,
|
| 2601 |
+
"trial_params": null
|
| 2602 |
+
}
|
checkpoint-18000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b89574d4fb532cbff16922a1304201180637df9c21820d655203050012091374
|
| 3 |
+
size 5329
|