Upload sft/d12/meta_002524.json with huggingface_hub
Browse files- sft/d12/meta_002524.json +82 -0
sft/d12/meta_002524.json
ADDED
|
@@ -0,0 +1,82 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"step": 2524,
|
| 3 |
+
"val_bpb": 0.07659919624310406,
|
| 4 |
+
"model_config": {
|
| 5 |
+
"sequence_len": 6144,
|
| 6 |
+
"vocab_size": 98304,
|
| 7 |
+
"n_layer": 12,
|
| 8 |
+
"n_head": 12,
|
| 9 |
+
"n_kv_head": 12,
|
| 10 |
+
"n_embd": 1536,
|
| 11 |
+
"window_pattern": "L",
|
| 12 |
+
"ve_layers": -1,
|
| 13 |
+
"grad_checkpoint": false
|
| 14 |
+
},
|
| 15 |
+
"user_config": {
|
| 16 |
+
"run": "mesosfer-sft-6k-mi300x",
|
| 17 |
+
"device_type": "",
|
| 18 |
+
"checkpoint_source": "sft",
|
| 19 |
+
"model_tag": null,
|
| 20 |
+
"model_step": null,
|
| 21 |
+
"load_optimizer": 1,
|
| 22 |
+
"num_iterations": -1,
|
| 23 |
+
"max_seq_len": 6144,
|
| 24 |
+
"device_batch_size": 16,
|
| 25 |
+
"total_batch_size": 491520,
|
| 26 |
+
"embedding_lr": null,
|
| 27 |
+
"unembedding_lr": null,
|
| 28 |
+
"matrix_lr": null,
|
| 29 |
+
"init_lr_frac": 0.8,
|
| 30 |
+
"warmup_ratio": 0.0,
|
| 31 |
+
"warmdown_ratio": 0.5,
|
| 32 |
+
"final_lr_frac": 0.0,
|
| 33 |
+
"eval_every": 200,
|
| 34 |
+
"eval_tokens": 20971520,
|
| 35 |
+
"chatcore_every": 200,
|
| 36 |
+
"chatcore_max_cat": -1,
|
| 37 |
+
"chatcore_max_sample": 24,
|
| 38 |
+
"chatcore_tasks": null,
|
| 39 |
+
"mmlu_epochs": 3,
|
| 40 |
+
"gsm8k_epochs": 4,
|
| 41 |
+
"cyber_defensive_epochs": 8,
|
| 42 |
+
"cloud_security_epochs": 8,
|
| 43 |
+
"multi_turn_soc_epochs": 0,
|
| 44 |
+
"tool_oriented_epochs": 8,
|
| 45 |
+
"mythos_epochs": 8,
|
| 46 |
+
"mythos_tool_calling_epochs": 4,
|
| 47 |
+
"mesosfer_validation_epochs": 0,
|
| 48 |
+
"gemini_teacher_epochs": 2,
|
| 49 |
+
"primus_instruct_epochs": 1,
|
| 50 |
+
"primus_reasoning_epochs": 1,
|
| 51 |
+
"cybernative_vuln_epochs": 3,
|
| 52 |
+
"openhermes_epochs": 1,
|
| 53 |
+
"ultrachat_epochs": 1,
|
| 54 |
+
"trendyol_cyber_epochs": 1,
|
| 55 |
+
"tiamz_cybersec_epochs": 2,
|
| 56 |
+
"alpaca_indonesian_epochs": 1,
|
| 57 |
+
"alpaca_gpt4_indonesian_epochs": 1,
|
| 58 |
+
"magpie_reasoning_epochs": 1,
|
| 59 |
+
"open_thoughts_epochs": 1,
|
| 60 |
+
"hermes_func_calling_epochs": 2,
|
| 61 |
+
"hermes_glaive_func_calling_epochs": 1,
|
| 62 |
+
"kimi_k3_traces_epochs": 4,
|
| 63 |
+
"cot_collection_epochs": 1,
|
| 64 |
+
"qyrou_reasoning_epochs": 1,
|
| 65 |
+
"nist_cybersec_epochs": 1,
|
| 66 |
+
"fenrir_v2_epochs": 1,
|
| 67 |
+
"code_feedback_epochs": 1,
|
| 68 |
+
"numinamath_cot_epochs": 1,
|
| 69 |
+
"aquilax_security_reasoning_epochs": 2,
|
| 70 |
+
"xlam_function_calling_epochs": 1,
|
| 71 |
+
"include_english_sft": 1,
|
| 72 |
+
"disable_cybersec_sft": false,
|
| 73 |
+
"rules_epochs": 4,
|
| 74 |
+
"tool_calling_epochs": 8,
|
| 75 |
+
"instruction_following_epochs": 4,
|
| 76 |
+
"instruction_polish_only": false,
|
| 77 |
+
"safety_artifact_epochs": 4,
|
| 78 |
+
"safety_artifact_only": false,
|
| 79 |
+
"local_sft_only": true,
|
| 80 |
+
"save_every": 200
|
| 81 |
+
}
|
| 82 |
+
}
|