KasparZ/mtext-250525_mistral-7B-v0.3_merged

Uploaded model

• —Developed by: KasparZ
• —License: apache-2.0
• —Finetuned from model : unsloth/mistral-7b-v0.3-bnb-4bit
• —maxseqlength = 4096
• —tokenizer.padtoken = tokenizer.eostoken
• —model.config.padtokenid = tokenizer.padtokenid
• —new_tokens = ["<|s|>", "<|e|>"]

• —LoRA
• —r = 128,
• —targetmodules = ["qproj", "kproj", "vproj", "oproj","gateproj", "upproj", "downproj","embedtokens", "lmhead"]
• —lora_alpha = 32,
• —lora_dropout = 0,
• —bias = "none",
• —usegradientcheckpointing = "unsloth",
• —random_state = 3407,
• —use_rslora = True,
• —loftq_config = None,

• —Training
• —perdevicetrainbatchsize = 1,
• —gradientaccumulationsteps = 8,
• —warmup_ratio = 0.1,
• —numtrainepochs = 1,
• —learning_rate = 1e-4,
• —embeddinglearningrate = 5e-5,
• —fp16 = True,
• —bf16 = False,
• —logging_steps = 1,
• —optim = "adamw_8bit",
• —weight_decay = 0.01,
• —lrschedulertype = "cosine",
• —seed = 3407,
• —output_dir = "outputs",
• —save_strategy = "steps",
• —save_steps = 50,
• —report_to = "none",

This mistral model was trained 2x faster with Unsloth and Huggingface's TRL library.

<img src="https://raw.githubusercontent.com/unslothai/unsloth/main/images/unsloth%20made%20with%20love.png" width="200"/>