reports/run/resolved.json

{
  "model": {
    "architectures": [
      "BalancedLM"
    ],
    "attention_bias": false,
    "attention_dropout": 0.0,
    "bos_token_id": 0,
    "dtype": "float32",
    "eos_token_id": 0,
    "head_dim": 64,
    "hidden_act": "silu",
    "hidden_size": 576,
    "initializer_range": 0.02,
    "intermediate_size": 2496,
    "layer_types": [
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention"
    ],
    "max_position_embeddings": 2048,
    "max_window_layers": 28,
    "model_type": "slayer_balanced",
    "num_attention_heads": 9,
    "num_hidden_layers": 26,
    "num_key_value_heads": 3,
    "pad_token_id": null,
    "rms_norm_eps": 1e-06,
    "rope_parameters": {
      "rope_theta": 10000.0,
      "rope_type": "default"
    },
    "sliding_window": null,
    "tie_word_embeddings": true,
    "transformers_version": "5.16.1",
    "use_cache": true,
    "use_sliding_window": false,
    "vocab_size": 24576
  },
  "train": {
    "seed": 1337,
    "block": 2048,
    "micro_batch": 16,
    "accumulation": 2,
    "adam_lr": 0.0015,
    "muon_lr": 0.02,
    "weight_decay": 0.1,
    "clip": 1.0,
    "checkpoint_every": 500,
    "log_every": 20,
    "compile": true
  },
  "parameters": 149333081,
  "world_size": 8,
  "tokens_per_update": 524288,
  "total_steps": 59368,
  "total_tokens": 31125929984,
  "warmup_steps": 2000,
  "initialization": "from_scratch"
}