mcloh/decision-models-benchmark
1
0
Fork 0
mirror of https://github.com/mcloh/decision-models-benchmark.git synced 2026-11-06 06:14:11 +00:00
Code Issues Packages Projects Releases Wiki Activity

Executor de avaliação, métricas e .gitignore para metadados de modelos

Browse Source 
- harness/evaluate.py: partição congelada com SHA-256, amostra com seed,
  lote 1, truncamento comum, contagem de tokens, latência (sem aquecimento),
  pico de RAM e GPU; teste bloqueado até o congelamento da configuração.
- analysis/metrics.py: acurácia, macro F1, Brier, NLL, ECE, acurácia seletiva.
- Diagnósticos por resposta (metadata.diagnostics); Laya informa opções
  colapsadas pelo orçamento de cabeçalho; head_max_len via DMB_LAYA_HEAD_MAX_LEN.
- infra/ds/collect.py e gpu_queue --env para pilotos e rodadas.
- .gitignore: de models/ só metadados (cards, configs, licenças, índices);
  pesos, tokenizadores grandes, runtimes, fontes clonados e caches ignorados.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This commit is contained in:
Marcos Lohmann
2026-10-07 15:06:57 -03:00
parent 344d1582fe
commit 328ccd1c2e
26 changed files with 5369 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

104
models/semif/Qwen__Qwen3.5-4B/851bf6e806efd8d0a36b00ddf55e13ccb7b8cd0a/config.json Normal file
View File

@@ -0,0 +1,104 @@
{
"architectures": [
"Qwen3_5ForConditionalGeneration"
],
"image_token_id": 248056,
"model_type": "qwen3_5",
"text_config": {
"attention_bias": false,
"attention_dropout": 0.0,
"attn_output_gate": true,
"dtype": "bfloat16",
"eos_token_id": 248044,
"full_attention_interval": 4,
"head_dim": 256,
"hidden_act": "silu",
"hidden_size": 2560,
"initializer_range": 0.02,
"intermediate_size": 9216,
"layer_types": [
"linear_attention",
"linear_attention",
"linear_attention",
"full_attention",
"linear_attention",
"linear_attention",
"linear_attention",
"full_attention",
"linear_attention",
"linear_attention",
"linear_attention",
"full_attention",
"linear_attention",
"linear_attention",
"linear_attention",
"full_attention",
"linear_attention",
"linear_attention",
"linear_attention",
"full_attention",
"linear_attention",
"linear_attention",
"linear_attention",
"full_attention",
"linear_attention",
"linear_attention",
"linear_attention",
"full_attention",
"linear_attention",
"linear_attention",
"linear_attention",
"full_attention"
],
"linear_conv_kernel_dim": 4,
"linear_key_head_dim": 128,
"linear_num_key_heads": 16,
"linear_num_value_heads": 32,
"linear_value_head_dim": 128,
"max_position_embeddings": 262144,
"mlp_only_layers": [],
"model_type": "qwen3_5_text",
"mtp_num_hidden_layers": 1,
"mtp_use_dedicated_embeddings": false,
"num_attention_heads": 16,
"num_hidden_layers": 32,
"num_key_value_heads": 4,
"rms_norm_eps": 1e-06,
"tie_word_embeddings": true,
"use_cache": true,
"vocab_size": 248320,
"mamba_ssm_dtype": "float32",
"rope_parameters": {
"mrope_interleaved": true,
"mrope_section": [
11,
11,
10
],
"rope_type": "default",
"rope_theta": 10000000,
"partial_rotary_factor": 0.25
}
},
"tie_word_embeddings": true,
"transformers_version": "4.57.0.dev0",
"video_token_id": 248057,
"vision_config": {
"deepstack_visual_indexes": [],
"depth": 24,
"hidden_act": "gelu_pytorch_tanh",
"hidden_size": 1024,
"in_channels": 3,
"initializer_range": 0.02,
"intermediate_size": 4096,
"model_type": "qwen3_5",
"num_heads": 16,
"num_position_embeddings": 2304,
"out_hidden_size": 2560,
"patch_size": 16,
"spatial_merge_size": 2,
"temporal_patch_size": 2
},
"vision_end_token_id": 248054,
"vision_start_token_id": 248053
}