Create config.json
Browse files- config.json +203 -0
config.json
ADDED
|
@@ -0,0 +1,203 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"model_name": "Troviku-1.1",
|
| 3 |
+
"model_type": "CausalLM",
|
| 4 |
+
"organization": "OpenTrouter",
|
| 5 |
+
"version": "1.1.0",
|
| 6 |
+
"release_date": "2025-01-15",
|
| 7 |
+
|
| 8 |
+
"architecture": {
|
| 9 |
+
"model_type": "transformer",
|
| 10 |
+
"hidden_size": 4096,
|
| 11 |
+
"num_hidden_layers": 32,
|
| 12 |
+
"num_attention_heads": 32,
|
| 13 |
+
"num_key_value_heads": 8,
|
| 14 |
+
"intermediate_size": 14336,
|
| 15 |
+
"hidden_act": "silu",
|
| 16 |
+
"max_position_embeddings": 8192,
|
| 17 |
+
"rope_theta": 10000.0,
|
| 18 |
+
"rope_scaling": null,
|
| 19 |
+
"attention_bias": false,
|
| 20 |
+
"attention_dropout": 0.0,
|
| 21 |
+
"mlp_bias": false
|
| 22 |
+
},
|
| 23 |
+
|
| 24 |
+
"tokenizer": {
|
| 25 |
+
"tokenizer_type": "BPE",
|
| 26 |
+
"vocab_size": 32768,
|
| 27 |
+
"bos_token": "<|begin_of_text|>",
|
| 28 |
+
"eos_token": "<|end_of_text|>",
|
| 29 |
+
"pad_token": "<|pad|>",
|
| 30 |
+
"unk_token": "<|unknown|>",
|
| 31 |
+
"special_tokens": [
|
| 32 |
+
"<|begin_of_text|>",
|
| 33 |
+
"<|end_of_text|>",
|
| 34 |
+
"<|pad|>",
|
| 35 |
+
"<|unknown|>",
|
| 36 |
+
"<|code_start|>",
|
| 37 |
+
"<|code_end|>",
|
| 38 |
+
"<|comment_start|>",
|
| 39 |
+
"<|comment_end|>",
|
| 40 |
+
"<|python|>",
|
| 41 |
+
"<|javascript|>",
|
| 42 |
+
"<|java|>",
|
| 43 |
+
"<|cpp|>",
|
| 44 |
+
"<|rust|>",
|
| 45 |
+
"<|go|>",
|
| 46 |
+
"<|typescript|>"
|
| 47 |
+
]
|
| 48 |
+
},
|
| 49 |
+
|
| 50 |
+
"training": {
|
| 51 |
+
"dataset_size_tokens": 500000000000,
|
| 52 |
+
"training_steps": 1000000,
|
| 53 |
+
"batch_size": 4096,
|
| 54 |
+
"learning_rate": 3e-4,
|
| 55 |
+
"warmup_steps": 2000,
|
| 56 |
+
"weight_decay": 0.1,
|
| 57 |
+
"gradient_clipping": 1.0,
|
| 58 |
+
"optimizer": "AdamW",
|
| 59 |
+
"mixed_precision": "bf16"
|
| 60 |
+
},
|
| 61 |
+
|
| 62 |
+
"inference": {
|
| 63 |
+
"default_temperature": 0.7,
|
| 64 |
+
"default_top_p": 0.95,
|
| 65 |
+
"default_top_k": 50,
|
| 66 |
+
"default_max_tokens": 2048,
|
| 67 |
+
"repetition_penalty": 1.1,
|
| 68 |
+
"frequency_penalty": 0.0,
|
| 69 |
+
"presence_penalty": 0.0
|
| 70 |
+
},
|
| 71 |
+
|
| 72 |
+
"capabilities": {
|
| 73 |
+
"languages": [
|
| 74 |
+
"python",
|
| 75 |
+
"javascript",
|
| 76 |
+
"typescript",
|
| 77 |
+
"java",
|
| 78 |
+
"c",
|
| 79 |
+
"cpp",
|
| 80 |
+
"csharp",
|
| 81 |
+
"rust",
|
| 82 |
+
"go",
|
| 83 |
+
"ruby",
|
| 84 |
+
"php",
|
| 85 |
+
"swift",
|
| 86 |
+
"kotlin",
|
| 87 |
+
"scala",
|
| 88 |
+
"r",
|
| 89 |
+
"sql",
|
| 90 |
+
"html",
|
| 91 |
+
"css",
|
| 92 |
+
"bash",
|
| 93 |
+
"powershell",
|
| 94 |
+
"lua",
|
| 95 |
+
"perl",
|
| 96 |
+
"haskell",
|
| 97 |
+
"julia",
|
| 98 |
+
"matlab"
|
| 99 |
+
],
|
| 100 |
+
"tasks": [
|
| 101 |
+
"code_generation",
|
| 102 |
+
"code_completion",
|
| 103 |
+
"code_explanation",
|
| 104 |
+
"code_review",
|
| 105 |
+
"bug_detection",
|
| 106 |
+
"test_generation",
|
| 107 |
+
"documentation_generation",
|
| 108 |
+
"code_translation",
|
| 109 |
+
"refactoring",
|
| 110 |
+
"algorithm_design"
|
| 111 |
+
],
|
| 112 |
+
"specializations": [
|
| 113 |
+
"algorithms",
|
| 114 |
+
"data_structures",
|
| 115 |
+
"web_development",
|
| 116 |
+
"api_design",
|
| 117 |
+
"database_queries",
|
| 118 |
+
"devops_scripting",
|
| 119 |
+
"data_science",
|
| 120 |
+
"machine_learning_code"
|
| 121 |
+
]
|
| 122 |
+
},
|
| 123 |
+
|
| 124 |
+
"benchmarks": {
|
| 125 |
+
"humaneval": {
|
| 126 |
+
"score": 0.72,
|
| 127 |
+
"date": "2025-01-10",
|
| 128 |
+
"notes": "Pass@1 score on HumanEval benchmark"
|
| 129 |
+
},
|
| 130 |
+
"mbpp": {
|
| 131 |
+
"score": 0.68,
|
| 132 |
+
"date": "2025-01-10",
|
| 133 |
+
"notes": "Pass@1 score on MBPP benchmark"
|
| 134 |
+
},
|
| 135 |
+
"code_contests": {
|
| 136 |
+
"score": 0.45,
|
| 137 |
+
"date": "2025-01-10",
|
| 138 |
+
"notes": "Problems solved at first attempt"
|
| 139 |
+
}
|
| 140 |
+
},
|
| 141 |
+
|
| 142 |
+
"safety": {
|
| 143 |
+
"content_filtering": true,
|
| 144 |
+
"malicious_code_detection": true,
|
| 145 |
+
"license_compliance": true,
|
| 146 |
+
"pii_redaction": true,
|
| 147 |
+
"toxicity_filtering": true
|
| 148 |
+
},
|
| 149 |
+
|
| 150 |
+
"deployment": {
|
| 151 |
+
"supported_frameworks": [
|
| 152 |
+
"transformers",
|
| 153 |
+
"vllm",
|
| 154 |
+
"tensorrt-llm",
|
| 155 |
+
"ctranslate2"
|
| 156 |
+
],
|
| 157 |
+
"quantization_support": [
|
| 158 |
+
"int8",
|
| 159 |
+
"int4",
|
| 160 |
+
"fp16",
|
| 161 |
+
"bf16"
|
| 162 |
+
],
|
| 163 |
+
"hardware_requirements": {
|
| 164 |
+
"minimum_vram_gb": 16,
|
| 165 |
+
"recommended_vram_gb": 24,
|
| 166 |
+
"minimum_ram_gb": 32,
|
| 167 |
+
"recommended_ram_gb": 64
|
| 168 |
+
}
|
| 169 |
+
},
|
| 170 |
+
|
| 171 |
+
"api": {
|
| 172 |
+
"endpoint": "https://api.opentrouter.ai/v1/chat/completions",
|
| 173 |
+
"model_identifier": "OpenTrouter/Troviku-1.1",
|
| 174 |
+
"rate_limits": {
|
| 175 |
+
"requests_per_minute": 60,
|
| 176 |
+
"tokens_per_minute": 90000
|
| 177 |
+
},
|
| 178 |
+
"pricing": {
|
| 179 |
+
"input_tokens_per_million": 0.50,
|
| 180 |
+
"output_tokens_per_million": 1.50,
|
| 181 |
+
"currency": "USD"
|
| 182 |
+
}
|
| 183 |
+
},
|
| 184 |
+
|
| 185 |
+
"license": {
|
| 186 |
+
"type": "OpenTrouter Model License",
|
| 187 |
+
"commercial_use": true,
|
| 188 |
+
"attribution_required": true,
|
| 189 |
+
"modifications_allowed": false,
|
| 190 |
+
"redistribution_allowed": false
|
| 191 |
+
},
|
| 192 |
+
|
| 193 |
+
"metadata": {
|
| 194 |
+
"authors": [
|
| 195 |
+
"OpenTrouter Research Team"
|
| 196 |
+
],
|
| 197 |
+
"contact": "[email protected]",
|
| 198 |
+
"repository": "https://github.com/OpenTrouter/Troviku-1.1",
|
| 199 |
+
"documentation": "https://docs.opentrouter.ai/troviku",
|
| 200 |
+
"paper": "https://arxiv.org/abs/placeholder",
|
| 201 |
+
"demo": "https://demo.opentrouter.ai/troviku"
|
| 202 |
+
}
|
| 203 |
+
}
|