Fix config.json: restore model_type and tokenizer files from v12
Browse files- config.json +62 -0
config.json
CHANGED
|
@@ -1,4 +1,66 @@
|
|
| 1 |
{
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2 |
"transformers.js_config": {
|
| 3 |
"use_external_data_format": {
|
| 4 |
"model_q4.onnx": 2
|
|
|
|
| 1 |
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"NemotronHForCausalLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_bias": false,
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"attention_head_dim": 128,
|
| 8 |
+
"auto_map": {
|
| 9 |
+
"AutoConfig": "configuration_nemotron_h.NemotronHConfig",
|
| 10 |
+
"AutoModelForCausalLM": "modeling_nemotron_h.NemotronHForCausalLM"
|
| 11 |
+
},
|
| 12 |
+
"bos_token_id": 1,
|
| 13 |
+
"chunk_size": 256,
|
| 14 |
+
"conv_kernel": 4,
|
| 15 |
+
"dtype": "bfloat16",
|
| 16 |
+
"eos_token_id": 2,
|
| 17 |
+
"expand": 2,
|
| 18 |
+
"head_dim": 128,
|
| 19 |
+
"hidden_dropout": 0.0,
|
| 20 |
+
"hidden_size": 3136,
|
| 21 |
+
"hybrid_override_pattern": "M-M-M-MM-M-M*-M-M*-M-M-M*-M-M-MM*-MMM-M-M-",
|
| 22 |
+
"initializer_range": 0.02,
|
| 23 |
+
"intermediate_size": 12544,
|
| 24 |
+
"layer_norm_epsilon": 1e-05,
|
| 25 |
+
"mamba_head_dim": 80,
|
| 26 |
+
"mamba_hidden_act": "silu",
|
| 27 |
+
"mamba_num_heads": 96,
|
| 28 |
+
"mamba_proj_bias": false,
|
| 29 |
+
"max_position_embeddings": 262144,
|
| 30 |
+
"mlp_bias": false,
|
| 31 |
+
"mlp_hidden_act": "relu2",
|
| 32 |
+
"model_name": "models/NVIDIA-Nemotron-3-Nano-4B",
|
| 33 |
+
"model_type": "nemotron_h",
|
| 34 |
+
"n_groups": 8,
|
| 35 |
+
"num_attention_heads": 40,
|
| 36 |
+
"num_hidden_layers": 42,
|
| 37 |
+
"num_key_value_heads": 8,
|
| 38 |
+
"num_logits_to_keep": 1,
|
| 39 |
+
"pad_token_id": 999,
|
| 40 |
+
"rescale_prenorm_residual": true,
|
| 41 |
+
"residual_in_fp32": false,
|
| 42 |
+
"rms_norm_eps": 1e-05,
|
| 43 |
+
"sliding_window": null,
|
| 44 |
+
"ssm_state_size": 128,
|
| 45 |
+
"tie_word_embeddings": false,
|
| 46 |
+
"time_step_floor": 0.0001,
|
| 47 |
+
"time_step_limit": [
|
| 48 |
+
0.0,
|
| 49 |
+
{
|
| 50 |
+
"__float__": "Infinity"
|
| 51 |
+
}
|
| 52 |
+
],
|
| 53 |
+
"time_step_max": 0.1,
|
| 54 |
+
"time_step_min": 0.001,
|
| 55 |
+
"time_step_rank": 256,
|
| 56 |
+
"transformers_version": "5.3.0",
|
| 57 |
+
"unsloth_fixed": true,
|
| 58 |
+
"unsloth_version": "2026.3.17",
|
| 59 |
+
"use_bias": false,
|
| 60 |
+
"use_cache": true,
|
| 61 |
+
"use_conv_bias": true,
|
| 62 |
+
"use_mamba_kernels": true,
|
| 63 |
+
"vocab_size": 131072,
|
| 64 |
"transformers.js_config": {
|
| 65 |
"use_external_data_format": {
|
| 66 |
"model_q4.onnx": 2
|