Added LoRA

README.md

@@ -0,0 +1,65 @@
+---
+language:
+- fi
+library_name: peft
+base_model: HPLT/gpt-7b-nordic-prerelease
+license: apache-2.0
+---
+
+# Model Card for Alpaca-Finnish-V1-7B-LoRA
+
+LoRA trained in 4-bit using [HPLT/gpt-7b-nordic-prerelease](https://huggingface.co/HPLT/gpt-7b-nordic-prerelease/) as the base model. Dataset used with the LoRA is [pinzhenchen/alpaca-cleaned-fi](https://huggingface.co/datasets/pinzhenchen/alpaca-cleaned-fi/).
+
+It uses Alpaca format but with a translated instruction at the start:
+```
+{
+    "instruction,output": "Alla on ohje, jossa kuvataan tehtävä. Kirjoita vastaus, joka täyttää pyynnön asianmukaisesti.\n\n### Instruction:\n%instruction%\n\n### Response:\n%output%",
+    "instruction,input,output": "Alla on ohje, jossa kuvataan tehtävä ja joka on yhdistetty kontekstia lisäävään syötteeseen. Kirjoita vastaus, joka täyttää pyynnön asianmukaisesti.\n\n### Instruction:\n%instruction%\n\n### Input:\n%input%\n\n### Response:\n%output%"
+}
+
+```
+
+Using the following settings:
+```json
+{
+  "lora_name": "Alpaca-Finnish-v1",
+  "always_override": false,
+  "q_proj_en": true,
+  "v_proj_en": true,
+  "k_proj_en": false,
+  "o_proj_en": false,
+  "gate_proj_en": false,
+  "down_proj_en": false,
+  "up_proj_en": false,
+  "save_steps": 250.0,
+  "micro_batch_size": 4,
+  "batch_size": 128,
+  "epochs": 3.0,
+  "learning_rate": "3e-4",
+  "lr_scheduler_type": "linear",
+  "lora_rank": 256,
+  "lora_alpha": 512,
+  "lora_dropout": 0.05,
+  "cutoff_len": 384,
+  "dataset": "alpaca_data_cleaned.fi",
+  "eval_dataset": "None",
+  "format": "alpaca-format-finnish",
+  "eval_steps": 100.0,
+  "raw_text_file": "None",
+  "overlap_len": 128,
+  "newline_favor_len": 128,
+  "higher_rank_limit": false,
+  "warmup_steps": 100.0,
+  "optimizer": "adamw_torch",
+  "hard_cut_string": "\\n\\n\\n",
+  "train_only_after": "",
+  "stop_at_loss": 0,
+  "add_eos_token": false,
+  "min_chars": 0.0,
+  "report_to": "None"
+}
+```
+
+### Framework versions
+
+- PEFT 0.8.2

adapter_config.json

@@ -0,0 +1,27 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "models/HPLT_gpt-7b-nordic-prerelease",
+  "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": 512,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 256,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "q_proj",
+    "v_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "use_rslora": false
+}

adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e95761245b006688771ae764a2b4b777adf523e6b45f486b64b1a2a029e60cac
+size 536888520

training_log.json

@@ -0,0 +1,12 @@
+{
+  "base_model_name": "HPLT_gpt-7b-nordic-prerelease",
+  "base_model_class": "LlamaForCausalLM",
+  "base_loaded_in_4bit": true,
+  "base_loaded_in_8bit": false,
+  "projections": "q, v",
+  "loss": 1.5969,
+  "grad_norm": 0.3398808538913727,
+  "learning_rate": 0.00020165562913907284,
+  "epoch": 0.99,
+  "current_steps": 12799
+}

training_parameters.json

@@ -0,0 +1,37 @@
+{
+  "lora_name": "Alpaca-Finnish-v1",
+  "always_override": false,
+  "q_proj_en": true,
+  "v_proj_en": true,
+  "k_proj_en": false,
+  "o_proj_en": false,
+  "gate_proj_en": false,
+  "down_proj_en": false,
+  "up_proj_en": false,
+  "save_steps": 250.0,
+  "micro_batch_size": 4,
+  "batch_size": 128,
+  "epochs": 3.0,
+  "learning_rate": "3e-4",
+  "lr_scheduler_type": "linear",
+  "lora_rank": 256,
+  "lora_alpha": 512,
+  "lora_dropout": 0.05,
+  "cutoff_len": 384,
+  "dataset": "alpaca_data_cleaned.fi",
+  "eval_dataset": "None",
+  "format": "alpaca-format-finnish",
+  "eval_steps": 100.0,
+  "raw_text_file": "None",
+  "overlap_len": 128,
+  "newline_favor_len": 128,
+  "higher_rank_limit": false,
+  "warmup_steps": 100.0,
+  "optimizer": "adamw_torch",
+  "hard_cut_string": "\\n\\n\\n",
+  "train_only_after": "",
+  "stop_at_loss": 0,
+  "add_eos_token": false,
+  "min_chars": 0.0,
+  "report_to": "None"
+}

training_prompt.json

@@ -0,0 +1,5 @@
+{
+  "template_type": "dataset",
+  "template_1": "Alla on ohje, jossa kuvataan teht\u00e4v\u00e4. Kirjoita vastaus, joka t\u00e4ytt\u00e4\u00e4 pyynn\u00f6n asianmukaisesti.\n\n### Instruction:\n%instruction%\n\n### Response:\n%output%",
+  "template_2": "Alla on ohje, jossa kuvataan teht\u00e4v\u00e4 ja joka on yhdistetty kontekstia lis\u00e4\u00e4v\u00e4\u00e4n sy\u00f6tteeseen. Kirjoita vastaus, joka t\u00e4ytt\u00e4\u00e4 pyynn\u00f6n asianmukaisesti.\n\n### Instruction:\n%instruction%\n\n### Input:\n%input%\n\n### Response:\n%output%"
+}