config.json

{
  "_name_or_path": "/tmp/models/tmp79btt9z02ijwklx8",
  "architectures": [
    "OlmoeForCausalLM"
  ],
  "attention_bias": false,
  "attention_dropout": 0.0,
  "clip_qkv": null,
  "eos_token_id": 50279,
  "hidden_act": "silu",
  "hidden_size": 2048,
  "initializer_range": 0.02,
  "intermediate_size": 1024,
  "max_position_embeddings": 4096,
  "model_type": "olmoe",
  "norm_topk_prob": false,
  "num_attention_heads": 16,
  "num_experts": 64,
  "num_experts_per_tok": 8,
  "num_hidden_layers": 16,
  "num_key_value_heads": 16,
  "output_router_logits": false,
  "pad_token_id": 1,
  "quantization_config": {
    "_load_in_4bit": false,
    "_load_in_8bit": true,
    "bnb_4bit_compute_dtype": "bfloat16",
    "bnb_4bit_quant_storage": "uint8",
    "bnb_4bit_quant_type": "fp4",
    "bnb_4bit_use_double_quant": false,
    "llm_int8_enable_fp32_cpu_offload": false,
    "llm_int8_has_fp16_weight": false,
    "llm_int8_skip_modules": [
      "lm_head"
    ],
    "llm_int8_threshold": 6.0,
    "load_in_4bit": false,
    "load_in_8bit": true,
    "quant_method": "bitsandbytes"
  },
  "rms_norm_eps": 1e-05,
  "rope_scaling": null,
  "rope_theta": 10000.0,
  "router_aux_loss_coef": 0.01,
  "tie_word_embeddings": false,
  "torch_dtype": "bfloat16",
  "transformers_version": "4.48.2",
  "use_cache": true,
  "vocab_size": 50304
}