An early checkpoint ~ Contextually obedient in answering, training curriculum is currently at training for contextual-refusals when the information provided is not sufficent to answer

Codestral-22B_Lora/adapter_config.json

@@ -0,0 +1,32 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "models\\Codestral-22B-Without-Mistral-Inference-Requirement",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 64,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 32,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "v_proj",
+    "k_proj",
+    "up_proj",
+    "gate_proj",
+    "q_proj",
+    "down_proj",
+    "o_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "use_rslora": false
+}

Codestral-22B_Lora/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d1d5db02575bfc83e2874cb6f78595de223de1d752843aed303f433742f3aaf8
+size 763469352

Codestral-22B_Lora/training_parameters.json

@@ -0,0 +1,37 @@
+{
+  "lora_name": "Contextual-Codestral",
+  "always_override": true,
+  "q_proj_en": true,
+  "v_proj_en": true,
+  "k_proj_en": true,
+  "o_proj_en": true,
+  "gate_proj_en": true,
+  "down_proj_en": true,
+  "up_proj_en": true,
+  "save_steps": 0,
+  "micro_batch_size": 42,
+  "batch_size": 24,
+  "epochs": 1.42,
+  "learning_rate": "6e-7",
+  "lr_scheduler_type": "polynomial",
+  "lora_rank": 8,
+  "lora_alpha": 16,
+  "lora_dropout": 0.05,
+  "cutoff_len": 8192,
+  "dataset": "0_Contextual",
+  "eval_dataset": "1_I_Cannot_Answer_That_For_You_From_Provided_Context",
+  "format": "Mermaid-format",
+  "eval_steps": 100,
+  "raw_text_file": "None",
+  "overlap_len": 128,
+  "newline_favor_len": 128,
+  "higher_rank_limit": false,
+  "warmup_steps": 1,
+  "optimizer": "adagrad",
+  "hard_cut_string": "\\n\\n\\n",
+  "train_only_after": "",
+  "stop_at_loss": 0.2,
+  "add_eos_token": true,
+  "min_chars": 0,
+  "report_to": "None"
+}

Codestral-22B_Lora/training_prompt.json

@@ -0,0 +1,6 @@
+{
+  "template_type": "dataset",
+  "template_1": "Code-Assistant-Request: %instruction%\n\n### Code:\n%output%",
+  "template_2": "Mermaid-Generator-Request: %instruction%\n\n### Input:\n%input%\n\n### Mermaid:\n%output%",
+  "template_3": "Contextual-Request: %prompt%\n\n### Contextual Response:\n%chosen%"
+}