Create generation_config.json

generation_config.json

@@ -0,0 +1,91 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 128000,
+  "eos_token_id": 128009,
+  "pad_token_id": 128001,
+  "transformers_version": "4.40.0",
+  "model_type": "llama",
+  "architectures": [
+    "LlamaForCausalLM"
+  ],
+  "do_sample": true,
+  "temperature": 0.7,
+  "top_p": 0.9,
+  "top_k": 50,
+  "repetition_penalty": 1.1,
+  "max_length": 131072,
+  "max_new_tokens": 4096,
+  "min_length": 0,
+  "min_new_tokens": null,
+  "early_stopping": false,
+  "num_beams": 1,
+  "num_beam_groups": 1,
+  "diversity_penalty": 0.0,
+  "length_penalty": 1.0,
+  "no_repeat_ngram_size": 0,
+  "encoder_no_repeat_ngram_size": 0,
+  "bad_words_ids": null,
+  "forced_bos_token_id": null,
+  "forced_eos_token_id": null,
+  "remove_invalid_values": false,
+  "exponential_decay_length_penalty": null,
+  "suppress_tokens": null,
+  "begin_suppress_tokens": null,
+  "forced_decoder_ids": null,
+  "num_return_sequences": 1,
+  "output_attentions": false,
+  "output_hidden_states": false,
+  "output_scores": false,
+  "return_dict_in_generate": true,
+  "use_cache": true,
+  "typical_p": 1.0,
+  "epsilon_cutoff": 0.0,
+  "eta_cutoff": 0.0,
+  "renormalize_logits": false,
+  "constraints": null,
+  "guidance_scale": null,
+  "low_memory": null,
+  "watermarking_config": null,
+  "presets": {
+    "creative": {
+      "temperature": 0.9,
+      "top_p": 0.95,
+      "top_k": 50,
+      "repetition_penalty": 1.1,
+      "description": "High creativity for stories, brainstorming, creative writing"
+    },
+    "precise": {
+      "temperature": 0.3,
+      "top_p": 0.85,
+      "top_k": 40,
+      "repetition_penalty": 1.15,
+      "description": "Low randomness for factual, technical, or code generation"
+    },
+    "balanced": {
+      "temperature": 0.7,
+      "top_p": 0.9,
+      "top_k": 50,
+      "repetition_penalty": 1.1,
+      "description": "Balanced for general purpose conversations"
+    },
+    "deterministic": {
+      "temperature": 0.0,
+      "top_p": 1.0,
+      "top_k": 1,
+      "repetition_penalty": 1.0,
+      "do_sample": false,
+      "description": "Fully deterministic output, same input = same output"
+    }
+  },
+  "stop_sequences": [
+    "<|im_end|>",
+    "<|endoftext|>",
+    "</s>"
+  ],
+  "chat_format": {
+    "system": "<|im_start|>system\n{content}<|im_end|>\n",
+    "user": "<|im_start|>user\n{content}<|im_end|>\n",
+    "assistant": "<|im_start|>assistant\n{content}<|im_end|>\n",
+    "prompt_suffix": "<|im_start|>assistant\n"
+  }
+}