Added LoRA

README.md

@@ -1,3 +1,65 @@
 ---
+language:
+- fi
+library_name: peft
+base_model: mpasila/gpt3-finnish-8B-gptq-4bit
 license: apache-2.0
 ---
+
+# Model Card for Model ID
+
+LoRA trained using [mpasila/gpt3-finnish-8B-gptq-4bit](https://huggingface.co/mpasila/gpt3-finnish-8B-gptq-4bit/) as the base model. Also the quantized model is based on this [TurkuNLP/gpt3-finnish-8B](https://huggingface.co/TurkuNLP/gpt3-finnish-8B/). Dataset used with the LoRA is [Finnish-NLP/Capybara-fi-deepl-translated-sft](https://huggingface.co/datasets/Finnish-NLP/Capybara-fi-deepl-translated-sft/) with some modifications so it uses Alpaca formatting [modified dataset](https://huggingface.co/datasets/mpasila/Capybara-fi-deepl-translated-sft-alpaca/).
+
+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": "Capybara_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": 32,
+  "lora_alpha": 64,
+  "lora_dropout": 0.05,
+  "cutoff_len": 256,
+  "dataset": "capybara_finnish_v1.1",
+  "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 +1,26 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "models/mpasila_gpt3-finnish-8B-gptq-4bit",
+  "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": [
+    "query_key_value"
+  ],
+  "task_type": "CAUSAL_LM",
+  "use_rslora": false
+}

adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7bd6fcdefd4588da08bb6030c6b9789195f0d5d2382e97c6b3ad4f9fe63df68b
+size 67118496

training_log.json

@@ -0,0 +1,17 @@
+{
+  "base_model_name": "mpasila_gpt3-finnish-8B-gptq-4bit",
+  "base_model_class": "BloomForCausalLM",
+  "base_loaded_in_4bit": false,
+  "base_loaded_in_8bit": false,
+  "projections": "query_key_value",
+  "loss": 2.3115,
+  "grad_norm": 0.10131204128265381,
+  "learning_rate": 3.1034482758620685e-05,
+  "epoch": 2.98,
+  "current_steps": 1051,
+  "train_runtime": 4886.022,
+  "train_samples_per_second": 0.869,
+  "train_steps_per_second": 0.007,
+  "total_flos": 112306365136896.0,
+  "train_loss": 2.4886908531188965
+}

training_parameters.json

@@ -0,0 +1,37 @@
+{
+  "lora_name": "Capybara_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": 32,
+  "lora_alpha": 64,
+  "lora_dropout": 0.05,
+  "cutoff_len": 256,
+  "dataset": "capybara_finnish_v1.1",
+  "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%"
+}