commit a4e10414482497907e8cd17d11d60e9ec14fd9ce Author: ModelHub XC Date: Thu May 7 10:37:46 2026 +0800 初始化项目，由ModelHub XC社区提供模型 Model: unsloth/granite-4.0-micro-GGUF Source: Original Platform diff --git a/.gitattributes b/.gitattributes new file mode 100644 index 0000000..df7e852 --- /dev/null +++ b/.gitattributes @@ -0,0 +1,62 @@ +*.7z filter=lfs diff=lfs merge=lfs -text +*.arrow filter=lfs diff=lfs merge=lfs -text +*.bin filter=lfs diff=lfs merge=lfs -text +*.bz2 filter=lfs diff=lfs merge=lfs -text +*.ckpt filter=lfs diff=lfs merge=lfs -text +*.ftz filter=lfs diff=lfs merge=lfs -text +*.gz filter=lfs diff=lfs merge=lfs -text +*.h5 filter=lfs diff=lfs merge=lfs -text +*.joblib filter=lfs diff=lfs merge=lfs -text +*.lfs.* filter=lfs diff=lfs merge=lfs -text +*.mlmodel filter=lfs diff=lfs merge=lfs -text +*.model filter=lfs diff=lfs merge=lfs -text +*.msgpack filter=lfs diff=lfs merge=lfs -text +*.npy filter=lfs diff=lfs merge=lfs -text +*.npz filter=lfs diff=lfs merge=lfs -text +*.onnx filter=lfs diff=lfs merge=lfs -text +*.ot filter=lfs diff=lfs merge=lfs -text +*.parquet filter=lfs diff=lfs merge=lfs -text +*.pb filter=lfs diff=lfs merge=lfs -text +*.pickle filter=lfs diff=lfs merge=lfs -text +*.pkl filter=lfs diff=lfs merge=lfs -text +*.pt filter=lfs diff=lfs merge=lfs -text +*.pth filter=lfs diff=lfs merge=lfs -text +*.rar filter=lfs diff=lfs merge=lfs -text +*.safetensors filter=lfs diff=lfs merge=lfs -text +saved_model/**/* filter=lfs diff=lfs merge=lfs -text +*.tar.* filter=lfs diff=lfs merge=lfs -text +*.tar filter=lfs diff=lfs merge=lfs -text +*.tflite filter=lfs diff=lfs merge=lfs -text +*.tgz filter=lfs diff=lfs merge=lfs -text +*.wasm filter=lfs diff=lfs merge=lfs -text +*.xz filter=lfs diff=lfs merge=lfs -text +*.zip filter=lfs diff=lfs merge=lfs -text +*.zst filter=lfs diff=lfs merge=lfs -text +*tfevents* filter=lfs diff=lfs merge=lfs -text +granite-4.0-micro-UD-IQ1_S.gguf filter=lfs diff=lfs merge=lfs -text +granite-4.0-micro-UD-Q3_K_XL.gguf filter=lfs diff=lfs merge=lfs -text +granite-4.0-micro-UD-Q2_K_XL.gguf filter=lfs diff=lfs merge=lfs -text +granite-4.0-micro-UD-IQ1_M.gguf filter=lfs diff=lfs merge=lfs -text +granite-4.0-micro-Q4_K_M.gguf filter=lfs diff=lfs merge=lfs -text +granite-4.0-micro-UD-IQ2_XXS.gguf filter=lfs diff=lfs merge=lfs -text +granite-4.0-micro-Q3_K_M.gguf filter=lfs diff=lfs merge=lfs -text +granite-4.0-micro-UD-IQ3_XXS.gguf filter=lfs diff=lfs merge=lfs -text +granite-4.0-micro-Q2_K.gguf filter=lfs diff=lfs merge=lfs -text +granite-4.0-micro-Q6_K.gguf filter=lfs diff=lfs merge=lfs -text +granite-4.0-micro-Q2_K_L.gguf filter=lfs diff=lfs merge=lfs -text +granite-4.0-micro-Q5_K_M.gguf filter=lfs diff=lfs merge=lfs -text +granite-4.0-micro-Q4_0.gguf filter=lfs diff=lfs merge=lfs -text +granite-4.0-micro-Q4_1.gguf filter=lfs diff=lfs merge=lfs -text +granite-4.0-micro-IQ4_NL.gguf filter=lfs diff=lfs merge=lfs -text +granite-4.0-micro-Q3_K_S.gguf filter=lfs diff=lfs merge=lfs -text +granite-4.0-micro-Q4_K_S.gguf filter=lfs diff=lfs merge=lfs -text +granite-4.0-micro-UD-Q6_K_XL.gguf filter=lfs diff=lfs merge=lfs -text +granite-4.0-micro-Q5_K_S.gguf filter=lfs diff=lfs merge=lfs -text +granite-4.0-micro-Q8_0.gguf filter=lfs diff=lfs merge=lfs -text +granite-4.0-micro-UD-Q8_K_XL.gguf filter=lfs diff=lfs merge=lfs -text +imatrix_unsloth.gguf_file filter=lfs diff=lfs merge=lfs -text +granite-4.0-micro-BF16.gguf filter=lfs diff=lfs merge=lfs -text +granite-4.0-micro-UD-Q5_K_XL.gguf filter=lfs diff=lfs merge=lfs -text +granite-4.0-micro-UD-Q4_K_XL.gguf filter=lfs diff=lfs merge=lfs -text +granite-4.0-micro-UD-IQ2_M.gguf filter=lfs diff=lfs merge=lfs -text +granite-4.0-micro-IQ4_XS.gguf filter=lfs diff=lfs merge=lfs -text diff --git a/README.md b/README.md new file mode 100644 index 0000000..7ed6951 --- /dev/null +++ b/README.md @@ -0,0 +1,650 @@ +--- +base_model: +- ibm-granite/granite-4.0-micro +license: apache-2.0 +library_name: transformers +tags: +- language +- unsloth +- granite-4.0 +--- +
+

+ See our collection for all versions of Granite-4.0 including GGUF, 4-bit & 16-bit formats. +

+

+ Learn to run Granite 4.0 correctly - Read our Guide. +

+

+ See Unsloth Dynamic 2.0 GGUFs for our quantization benchmarks. +

+
+ + + + + + + + + +
+

✨ Read our Granite-4.0 Guide here!

+
+ +- Fine-tune Granite-4.0 for free using our [Google Colab notebook](https://colab.research.google.com/github/unslothai/notebooks/blob/main/nb/Granite4.0.ipynb) +- Read our Blog about Granite-4.0 support: https://docs.unsloth.ai/new/ibm-granite-4.0 +- View the rest of our notebooks in our [docs here](https://docs.unsloth.ai/get-started/unsloth-notebooks). + +# Granite-4.0-Micro + +**Model Summary:** +Granite-4.0-Micro is a 3B parameter long-context instruct model finetuned from *Granite-4.0-Micro-Base* using a combination of open source instruction datasets with permissive license and internally collected synthetic datasets. This model is developed using a diverse set of techniques with a structured chat format, including supervised finetuning, model alignment using reinforcement learning, and model merging. Granite 4.0 instruct models feature improved *instruction following (IF)* and *tool-calling* capabilities, making them more effective in enterprise applications. + +- **Developers:** Granite Team, IBM +- **HF Collection:** [Granite 4.0 Language Models HF Collection](https://huggingface.co/collections/ibm-granite/granite-40-language-models-6811a18b820ef362d9e5a82c) +- **GitHub Repository:** [ibm-granite/granite-4.0-language-models](https://github.com/ibm-granite/granite-4.0-language-models) +- **Website**: [Granite Docs](https://www.ibm.com/granite/docs/) +- **Release Date**: October 2nd, 2025 +- **License:** [Apache 2.0](https://www.apache.org/licenses/LICENSE-2.0) + +**Supported Languages:** +English, German, Spanish, French, Japanese, Portuguese, Arabic, Czech, Italian, Korean, Dutch, and Chinese. Users may finetune Granite 4.0 models for languages beyond these languages. + +**Intended use:** +The model is designed to follow general instructions and can serve as the foundation for AI assistants across diverse domains, including business applications, as well as for LLM agents equipped with tool-use capabilities. + +*Capabilities* +* Summarization +* Text classification +* Text extraction +* Question-answering +* Retrieval Augmented Generation (RAG) +* Code related tasks +* Function-calling tasks +* Multilingual dialog use cases +* Fill-In-the-Middle (FIM) code completions + + + +**Generation:** +This is a simple example of how to use Granite-4.0-Micro model. + +Install the following libraries: + +```shell +pip install torch torchvision torchaudio +pip install accelerate +pip install transformers +``` +Then, copy the snippet from the section that is relevant for your use case. + +```python +import torch +from transformers import AutoModelForCausalLM, AutoTokenizer + +device = "cuda" +model_path = "ibm-granite/granite-4.0-micro" +tokenizer = AutoTokenizer.from_pretrained(model_path) +# drop device_map if running on CPU +model = AutoModelForCausalLM.from_pretrained(model_path, device_map=device) +model.eval() +# change input text as desired +chat = [ + { "role": "user", "content": "Please list one IBM Research laboratory located in the United States. You should only output its name and location." }, +] +chat = tokenizer.apply_chat_template(chat, tokenize=False, add_generation_prompt=True) +# tokenize the text +input_tokens = tokenizer(chat, return_tensors="pt").to(device) +# generate output tokens +output = model.generate(**input_tokens, + max_new_tokens=100) +# decode output tokens into text +output = tokenizer.batch_decode(output) +# print output +print(output[0]) +``` + +Expected output: +```shell +<|start_of_role|>user<|end_of_role|>Please list one IBM Research laboratory located in the United States. You should only output its name and location.<|end_of_text|> +<|start_of_role|>assistant<|end_of_role|>Almaden Research Center, San Jose, California<|end_of_text|> +``` + +**Tool-calling:** +Granite-4.0-Micro comes with enhanced tool calling capabilities, enabling seamless integration with external functions and APIs. To define a list of tools please follow OpenAI's function [definition schema](https://platform.openai.com/docs/guides/function-calling?api-mode=responses#defining-functions). + +This is an example of how to use Granite-4.0-Micro model tool-calling ability: + +```python +import torch +from transformers import AutoModelForCausalLM, AutoTokenizer + +device = "cuda" +model_path = "ibm-granite/granite-4.0-micro" +tokenizer = AutoTokenizer.from_pretrained(model_path) +# drop device_map if running on CPU +model = AutoModelForCausalLM.from_pretrained(model_path, device_map=device) +model.eval() + +tools = [ + { + "type": "function", + "function": { + "name": "get_current_weather", + "description": "Get the current weather for a specified city.", + "parameters": { + "type": "object", + "properties": { + "city": { + "type": "string", + "description": "Name of the city" + } + }, + "required": ["city"] + } + } + } +] + +# change input text as desired +chat = [ + { "role": "user", "content": "What's the weather like in Boston right now?" }, +] +chat = tokenizer.apply_chat_template(chat, \ + tokenize=False, \ + tools=tools, \ + add_generation_prompt=True) +# tokenize the text +input_tokens = tokenizer(chat, return_tensors="pt").to(device) +# generate output tokens +output = model.generate(**input_tokens, + max_new_tokens=100) +# decode output tokens into text +output = tokenizer.batch_decode(output) +# print output +print(output[0]) +``` + +Expected output: +```shell +<|start_of_role|>system<|end_of_role|>You are a helpful assistant with access to the following tools. You may call one or more tools to assist with the user query. + +You are provided with function signatures within XML tags: +- +- unsloth +{"type": "function", "function": {"name": "get_current_weather", "description": "Get the current weather for a specified city.", "parameters": {"type": "object", "properties": {"city": {"type": "string", "description": "Name of the city"}}, "required": ["city"]}}} + + +For each tool call, return a json object with function name and arguments within XML tags: +- +- unsloth +{"name": , "arguments": } +. If a tool does not exist in the provided list of tools, notify the user that you do not have the ability to fulfill the request.<|end_of_text|> +<|start_of_role|>user<|end_of_role|>What's the weather like in Boston right now?<|end_of_text|> +<|start_of_role|>assistant<|end_of_role|> +{"name": "get_current_weather", "arguments": {"city": "Boston"}} +<|end_of_text|> +``` + + + +**Evaluation Results:** + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
BenchmarksMetricMicro DenseH Micro DenseH Tiny MoEH Small MoE
+ General Tasks +
MMLU5-shot65.9867.4368.6578.44
MMLU-Pro5-shot, CoT44.543.4844.9455.47
BBH3-shot, CoT72.4869.3666.3481.62
AGI EVAL0-shot, CoT64.295962.1570.63
GPQA0-shot, CoT30.1432.1532.5940.63
+ Alignment Tasks +
AlpacaEval 2.029.4931.4930.6142.48
IFEvalInstruct, Strict85.586.9484.7889.87
IFEvalPrompt, Strict79.1281.7178.185.22
IFEvalAverage82.3184.3281.4487.55
ArenaHard25.8436.1535.7546.48
+ Math Tasks +
GSM8K8-shot85.4581.3584.6987.27
GSM8K Symbolic8-shot79.8277.581.187.38
Minerva Math0-shot, CoT62.0666.4469.6474
DeepMind Math0-shot, CoT44.5643.8349.9259.33
+ Code Tasks +
HumanEvalpass@180818388
HumanEval+pass@172757683
MBPPpass@172738084
MBPP+pass@164646971
CRUXEval-Opass@141.541.2539.6350.25
BigCodeBenchpass@139.2137.941.0646.23
+ Tool Calling Tasks +
BFCL v359.9857.5657.6564.69
+ Multilingual Tasks +
MULTIPLEpass@149.2149.4655.8357.37
MMMLU5-shot55.1455.1961.8769.69
INCLUDE5-shot51.6250.5153.1263.97
MGSM8-shot28.5644.4845.3638.72
+ Safety +
SALAD-Bench97.0696.2897.7797.3
AttaQ86.0584.4486.6186.64
+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
Multilingual Benchmarks and thr included languages:
Benchmarks# LangsLanguages
MMMLU11ar, de, en, es, fr, ja, ko, pt, zh, bn, hi
INCLUDE14hi, bn, ta, te, ar, de, es, fr, it, ja, ko, nl, pt, zh
MGSM5en, es, fr, ja, zh
+ +**Model Architecture:** + +Granite-4.0-Micro baseline is built on a decoder-only dense transformer architecture. Core components of this architecture are: GQA, RoPE, MLP with SwiGLU, RMSNorm, and shared input/output embeddings. + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
ModelMicro DenseH Micro DenseH Tiny MoEH Small MoE
Embedding size2560204815364096
Number of layers40 attention4 attention / 36 Mamba24 attention / 36 Mamba24 attention / 36 Mamba2
Attention head size6464128128
Number of attention heads40321232
Number of KV heads8848
Mamba2 state size-128128128
Number of Mamba2 heads-6448128
MLP / Shared expert hidden size8192819210241536
Num. Experts--6472
Num. active Experts--610
Expert hidden size--512768
MLP activationSwiGLUSwiGLUSwiGLUSwiGLU
Sequence length128K128K128K128K
Position embeddingRoPENoPENoPENoPE
# Parameters3B3B7B32B
# Active parameters3B3B1B9B
+ +**Training Data:** +Overall, our SFT data is largely comprised of three key sources: (1) publicly available datasets with permissive license, (2) internal synthetic data targeting specific capabilities, and (3) a select set of human-curated data. + +**Infrastructure:** +We trained the Granite 4.0 Language Models utilizing an NVIDIA GB200 NVL72 cluster hosted in CoreWeave. Intra-rack communication occurs via the 72-GPU NVLink domain, and a non-blocking, full Fat-Tree NDR 400 Gb/s InfiniBand network provides inter-rack communication. This cluster provides a scalable and efficient infrastructure for training our models over thousands of GPUs. + +**Ethical Considerations and Limitations:** +Granite 4.0 Instruction Models are primarily finetuned using instruction-response pairs mostly in English, but also multilingual data covering multiple languages. Although this model can handle multilingual dialog use cases, its performance might not be similar to English tasks. In such case, introducing a small number of examples (few-shot) can help the model in generating more accurate outputs. While this model has been aligned by keeping safety in consideration, the model may in some cases produce inaccurate, biased, or unsafe responses to user prompts. So we urge the community to use this model with proper safety testing and tuning tailored for their specific tasks. + +**Resources** +- ⭐️ Learn about the latest updates with Granite: https://www.ibm.com/granite +- 📄 Get started with tutorials, best practices, and prompt engineering advice: https://www.ibm.com/granite/docs/ +- 💡 Learn about the latest Granite learning resources: https://ibm.biz/granite-learning-resources + + \ No newline at end of file diff --git a/granite-4.0-micro-BF16.gguf b/granite-4.0-micro-BF16.gguf new file mode 100644 index 0000000..0bc95da --- /dev/null +++ b/granite-4.0-micro-BF16.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b6b27fbc5243e4294afcc527a6bfbe6809df1b665af0179952e4377e2ec278e5 +size 6809656800 diff --git a/granite-4.0-micro-IQ4_NL.gguf b/granite-4.0-micro-IQ4_NL.gguf new file mode 100644 index 0000000..baf07e1 --- /dev/null +++ b/granite-4.0-micro-IQ4_NL.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8faf8f8813bf9ea1c62b4284bb70e8ac0a4d687249c6932b660fc75956aa06a1 +size 1991163648 diff --git a/granite-4.0-micro-IQ4_XS.gguf b/granite-4.0-micro-IQ4_XS.gguf new file mode 100644 index 0000000..7488626 --- /dev/null +++ b/granite-4.0-micro-IQ4_XS.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cb48b3f1cfd4b3500b7007ec4fb9729e28224ea3f94aaec7892f3a101d040a32 +size 1894498048 diff --git a/granite-4.0-micro-Q2_K.gguf b/granite-4.0-micro-Q2_K.gguf new file mode 100644 index 0000000..7ac63fa --- /dev/null +++ b/granite-4.0-micro-Q2_K.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c1fdfcae12679ac41cf55875de3647804643552457bdffacee0d4311f226e609 +size 1371438848 diff --git a/granite-4.0-micro-Q2_K_L.gguf b/granite-4.0-micro-Q2_K_L.gguf new file mode 100644 index 0000000..7ac63fa --- /dev/null +++ b/granite-4.0-micro-Q2_K_L.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c1fdfcae12679ac41cf55875de3647804643552457bdffacee0d4311f226e609 +size 1371438848 diff --git a/granite-4.0-micro-Q3_K_M.gguf b/granite-4.0-micro-Q3_K_M.gguf new file mode 100644 index 0000000..8ea8626 --- /dev/null +++ b/granite-4.0-micro-Q3_K_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b167c94cefd31b763f00c16d0c5ff74c76d4b612caac27e65607a1d4f36b010a +size 1725579008 diff --git a/granite-4.0-micro-Q3_K_S.gguf b/granite-4.0-micro-Q3_K_S.gguf new file mode 100644 index 0000000..808f954 --- /dev/null +++ b/granite-4.0-micro-Q3_K_S.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1168270175756e39de329c311bc70e90a508390eb64271e83dc9dd874bd666a7 +size 1566818048 diff --git a/granite-4.0-micro-Q4_0.gguf b/granite-4.0-micro-Q4_0.gguf new file mode 100644 index 0000000..2bd5211 --- /dev/null +++ b/granite-4.0-micro-Q4_0.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c4763a66780871dc656a5b579cd23fe8a77efbe2d253b03a693efa713de2c371 +size 1991163648 diff --git a/granite-4.0-micro-Q4_1.gguf b/granite-4.0-micro-Q4_1.gguf new file mode 100644 index 0000000..4e8e5d2 --- /dev/null +++ b/granite-4.0-micro-Q4_1.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:93bec266d1e2dd8eaec18f8e2115d39276b7210bdf6e0c49ab1b2ee85bdacc9f +size 2181218048 diff --git a/granite-4.0-micro-Q4_K_M.gguf b/granite-4.0-micro-Q4_K_M.gguf new file mode 100644 index 0000000..be79549 --- /dev/null +++ b/granite-4.0-micro-Q4_K_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3ed0a9f397bab63eb93231449086ce5dc3f309598d007825fc01cfd32c4b93e1 +size 2099502848 diff --git a/granite-4.0-micro-Q4_K_S.gguf b/granite-4.0-micro-Q4_K_S.gguf new file mode 100644 index 0000000..ed72f16 --- /dev/null +++ b/granite-4.0-micro-Q4_K_S.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bc154bb604fc12af4f1682ab94e33afb2aeaa68b8df6afbd32a537cf02cc89e3 +size 1998372608 diff --git a/granite-4.0-micro-Q5_K_M.gguf b/granite-4.0-micro-Q5_K_M.gguf new file mode 100644 index 0000000..6d2536c --- /dev/null +++ b/granite-4.0-micro-Q5_K_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7f77222fa5fb8445b2fbcac13577f1f4d76896a304a88d27d6d5baa382a8970c +size 2437013248 diff --git a/granite-4.0-micro-Q5_K_S.gguf b/granite-4.0-micro-Q5_K_S.gguf new file mode 100644 index 0000000..be71f4d --- /dev/null +++ b/granite-4.0-micro-Q5_K_S.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:373157397937a931d139a484aa0326ebba0392127636aa669666058e8406e813 +size 2377826048 diff --git a/granite-4.0-micro-Q6_K.gguf b/granite-4.0-micro-Q6_K.gguf new file mode 100644 index 0000000..550b054 --- /dev/null +++ b/granite-4.0-micro-Q6_K.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f55d3414e2c8650c1a1d718d5e6cbf0b06b78f17668d047d28b69ab82445ba0d +size 2795618048 diff --git a/granite-4.0-micro-Q8_0.gguf b/granite-4.0-micro-Q8_0.gguf new file mode 100644 index 0000000..24f96f6 --- /dev/null +++ b/granite-4.0-micro-Q8_0.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9ece7079edab674f2192e5339e9198e5844ae348fb289982dc0da1ec9469b830 +size 3619692288 diff --git a/granite-4.0-micro-UD-IQ1_M.gguf b/granite-4.0-micro-UD-IQ1_M.gguf new file mode 100644 index 0000000..e69bc3d --- /dev/null +++ b/granite-4.0-micro-UD-IQ1_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c66f9f1bc3e31e319ebf42eb870440b606374c7d7ff163ff07868d9b943e9ea1 +size 938655488 diff --git a/granite-4.0-micro-UD-IQ1_S.gguf b/granite-4.0-micro-UD-IQ1_S.gguf new file mode 100644 index 0000000..3965f20 --- /dev/null +++ b/granite-4.0-micro-UD-IQ1_S.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bc0e541d4a5b16ecee62a75c1c9882912c84f74435e5bc85d96c236320f0808e +size 882294528 diff --git a/granite-4.0-micro-UD-IQ2_M.gguf b/granite-4.0-micro-UD-IQ2_M.gguf new file mode 100644 index 0000000..9433b60 --- /dev/null +++ b/granite-4.0-micro-UD-IQ2_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:40932e3a07b51848a355f4631c6057cbed24a198c377236ab3871eb1baa3011b +size 1265680128 diff --git a/granite-4.0-micro-UD-IQ2_XXS.gguf b/granite-4.0-micro-UD-IQ2_XXS.gguf new file mode 100644 index 0000000..1dc8035 --- /dev/null +++ b/granite-4.0-micro-UD-IQ2_XXS.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f225d235e05c000e4b767664a481158e7161bac3158626451d83c8406525980e +size 1035525888 diff --git a/granite-4.0-micro-UD-IQ3_XXS.gguf b/granite-4.0-micro-UD-IQ3_XXS.gguf new file mode 100644 index 0000000..de80089 --- /dev/null +++ b/granite-4.0-micro-UD-IQ3_XXS.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0a84765d194736dc489da43e4e77ed8c01dbb06e898a32f1b55b5945109d022f +size 1404083968 diff --git a/granite-4.0-micro-UD-Q2_K_XL.gguf b/granite-4.0-micro-UD-Q2_K_XL.gguf new file mode 100644 index 0000000..728319c --- /dev/null +++ b/granite-4.0-micro-UD-Q2_K_XL.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7af1ced1d091f14b029d57fe6b9709ba8686911e4cd6f980c884a1ef83e8d979 +size 1402363648 diff --git a/granite-4.0-micro-UD-Q3_K_XL.gguf b/granite-4.0-micro-UD-Q3_K_XL.gguf new file mode 100644 index 0000000..771e882 --- /dev/null +++ b/granite-4.0-micro-UD-Q3_K_XL.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d924a560af878eacba0e8340b25159e18e066500d026518fe3e50e7ff8918828 +size 1768689408 diff --git a/granite-4.0-micro-UD-Q4_K_XL.gguf b/granite-4.0-micro-UD-Q4_K_XL.gguf new file mode 100644 index 0000000..c9fa079 --- /dev/null +++ b/granite-4.0-micro-UD-Q4_K_XL.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:90849f9762f9a19f0d49b7822a01588dfef7218fc111bcbe291fee81a8431e8d +size 2137308928 diff --git a/granite-4.0-micro-UD-Q5_K_XL.gguf b/granite-4.0-micro-UD-Q5_K_XL.gguf new file mode 100644 index 0000000..b6f829c --- /dev/null +++ b/granite-4.0-micro-UD-Q5_K_XL.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2f6ade9828221e21a3d3149eaee1bd3a90c6a19dae415b990d74d7d3d72673ba +size 2449813248 diff --git a/granite-4.0-micro-UD-Q6_K_XL.gguf b/granite-4.0-micro-UD-Q6_K_XL.gguf new file mode 100644 index 0000000..9d5d907 --- /dev/null +++ b/granite-4.0-micro-UD-Q6_K_XL.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fbcb699f488d2b557aee6cb9003c76b7693a26905349f34d3b7139224c11f733 +size 3048300288 diff --git a/granite-4.0-micro-UD-Q8_K_XL.gguf b/granite-4.0-micro-UD-Q8_K_XL.gguf new file mode 100644 index 0000000..5fddd13 --- /dev/null +++ b/granite-4.0-micro-UD-Q8_K_XL.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b2ef94881daf3f3e27801394964eab9b0e7e2c92765101b3d85de7164103e8f8 +size 4241465088 diff --git a/imatrix_unsloth.gguf_file b/imatrix_unsloth.gguf_file new file mode 100644 index 0000000..d9e69d6 --- /dev/null +++ b/imatrix_unsloth.gguf_file @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2d90b81af88d25953d25417df32c44f7579f1fee5a3195260f8361c33ad4a17e +size 3811424 diff --git a/params b/params new file mode 100644 index 0000000..26b3b6e --- /dev/null +++ b/params @@ -0,0 +1,7 @@ +{ + "stop": [ + "<|end_of_text|>" + ], + "temperature": 0.0, + "min_p" : 0.00 +} \ No newline at end of file diff --git a/template b/template new file mode 100644 index 0000000..bd32be7 --- /dev/null +++ b/template @@ -0,0 +1,158 @@ +{{- /* + +------ MESSAGE PARSING ------ + +*/}} +{{- /* +Declare the system prompt chunks used for different features +*/}} +{{- $tools_system_message_prefix := "You are a helpful assistant with access to the following tools. You may call one or more tools to assist with the user query.\n\nYou are provided with function signatures within XML tags:\n" }} +{{- $tools_system_message_suffix := "\n\n\nFor each tool call, return a json object with function name and arguments within XML tags:\n\n{\"name\": , \"arguments\": }\n. If a tool does not exist in the provided list of tools, notify the user that you do not have the ability to fulfill the request." }} +{{- $documents_system_message_prefix := "You are a helpful assistant with access to the following documents. You may use one or more documents to assist with the user query.\n\nYou are given a list of documents within XML tags:\n" }} +{{- $documents_system_message_suffix := "\n\n\nWrite the response to the user's input by strictly aligning with the facts in the provided documents. If the information needed to answer the question is not available in the documents, inform the user that the question cannot be answered based on the available data." }} +{{- $default_system_message := "You are a helpful assistant. Please ensure responses are professional, accurate, and safe." }} + +{{- /* +Declare the prompt structure variables to be filled in from messages +*/}} +{{- $tools_system_message := "" }} +{{- $documents_system_message := "" }} +{{- $system_message := "" }} +{{- $document_counter := 0 }} +{{- $last_query_index := 0 }} + +{{- /* +Output parsing heuristic + +Ollama has very specific heuristics for parsing the tags for thinking and tool +calling. Rather than contorting the actual message expansion to match the +expectations, we explicitly put them here behind an unreachable condition. +*/}} +{{- if false }} + {{- /* tool calls: https://github.com/ollama/ollama/blob/main/tools/template.go#L17 */}} + {{- if .ToolCalls }}{{- end }} +{{- end }} + +{{- /* +Create tools system message chunk +*/}} +{{- if .Tools }} + {{- $tools_system_message = print $tools_system_message_prefix }} + {{- range $_, $tool_body := .Tools }} + {{- $tools_system_message = print $tools_system_message "\n" (json $tool_body) }} + {{- end }} + {{- $tools_system_message = print $tools_system_message $tools_system_message_suffix }} +{{- end }} + + +{{- /* +Loop over messages to parse variables: + +- User provided documents in the "document" role +- Last user query index +- Initial system message + +NOTE: Since Ollama collates consecutive roles, for documents, we work around + this by allowing the role to contain a qualifier after the role string. This + is also then used as the title of the document. +*/ -}} +{{- range $index, $_ := .Messages }} + {{- if (and (eq .Role "system") (eq $index 0)) }} + {{- if ne $system_message "" }} + {{- $system_message = print $system_message "\n" }} + {{- end }} + {{- $system_message = print $system_message .Content }} + {{- else if eq .Role "user" }} + {{- /* + NOTE: 31 == len ''. We only check the + prefix match since go template doesn't support negative indexing. + */}} + {{- if or (lt (len .Content) 31) (ne (slice .Content 0 15) "") }} + {{- $last_query_index = $index }} + {{- end }} + {{- else if (and (ge (len .Role) 8) (eq (slice .Role 0 8) "document")) }} + {{- if (eq $document_counter 0)}} + {{- $documents_system_message = print $documents_system_message_prefix}} + {{- end }} + {{- $identifier := ""}} + {{- if (ge (len .Role) 9) }} + {{- $identifier = slice .Role 9}} + {{- end }} + {{- if (eq $identifier "") }} + {{- $identifier := print $document_counter}} + {{- end }} + {{- $documents_system_message = print $documents_system_message "\n{\"doc_id\": " $document_counter ", \"title\": \"" $identifier "\", \"text\": \"" .Content "\"}"}} + {{- $document_counter = len (printf "a%*s" $document_counter "")}} + {{- end }} +{{- end }} +{{- if (ne $document_counter 0) }} +{{- $documents_system_message = print $documents_system_message $documents_system_message_suffix}} +{{- end }} + +{{- /* +Construct the full system message +*/}} +{{- if ne $tools_system_message "" }} + {{- if ne $system_message "" }} + {{- $system_message = print $system_message "\n\n" }} + {{- end }} + {{- $system_message = print $system_message $tools_system_message }} +{{- end }} +{{- if ne $documents_system_message "" }} + {{- if ne $system_message "" }} + {{- $system_message = print $system_message "\n\n" }} + {{- end }} + {{- $system_message = print $system_message $documents_system_message }} +{{- end }} + +{{- /* + +------ TEMPLATE EXPANSION ------ + +*/}} +{{- if ne $system_message "" -}} +<|start_of_role|>system<|end_of_role|>{{ $system_message }}<|end_of_text|>{{ "\n" }} +{{- else -}} +<|start_of_role|>system<|end_of_role|>{{ $default_system_message }}<|end_of_text|>{{ "\n" }} +{{- end }} +{{- $prev_role := "" }} +{{- range $message_index, $_ := .Messages }} + {{- $next_message_index := len (printf "a%*s" $message_index "") }} + {{- if or (eq .Role "user") (and (eq .Role "system") (ne $message_index 0)) }} + {{- "" }}<|start_of_role|>{{- .Role }}<|end_of_role|>{{- .Content }}<|end_of_text|>{{ "\n" }} + {{- else if eq .Role "assistant" -}} + {{- "" }}<|start_of_role|>{{ .Role }}<|end_of_role|>{{ .Content }} + + {{- /* Expand tool calls */}} + {{- $content := .Content }} + {{- if .ToolCalls }} + {{- range $tool_idx, $tool_call := .ToolCalls }} + {{- if or (ne $content "") (ne $tool_idx 0) }} + {{- "\n" }} + {{- end }} + {{- print "\n{\"name\": \"" $tool_call.Function.Name "\", \"arguments\": " (json $tool_call.Function.Arguments) "}\n" }} + {{- end }} + {{- end }} + + {{- /* End assistant block */}}<|end_of_text|>{{ "\n" }} + + {{- else if eq .Role "tool" }} + {{- if (ne $prev_role "tool") -}} + <|start_of_role>user<|end_of_role|> + {{- end }} + {{- "\n" }}{{ print "\n" .Content "\n" }} + {{- if ne $next_message_index (len $.Messages) }} + {{- $next_element := index $.Messages $next_message_index }} + {{- if ne $next_element.Role "tool" -}} + <|end_of_text|>{{ "\n" }} + {{- end }} + {{- else -}} + <|end_of_text|>{{ "\n" }} + {{- end }} + {{- end }} + + {{- /* If not an assistant message at the end, add generation prompt */}} + {{- if and (eq $next_message_index (len $.Messages)) (ne .Role "assistant") }} + {{- "<|start_of_role|>assistant<|end_of_role|>" }} + {{- end }} +{{- end }} \ No newline at end of file