Standalone merged BF16 model from epoch 3.970178926441352, step 1500. Dataset: CompassioninMachineLearning/urban12738cleaned at ef7c0e742df63ea319e35d02d9f6ba63d6e7c68d. Training: 10,072 distinct documents plus 2,000 repeat exposures per epoch; 200 disjoint validation documents. Merged with Unsloth's native savepretrainedmerged(savemethod="merged16bit"). Weights are validated BF16 and packaged losslessly into eight safetensors shards. No adapter is required to load this model. See runmanifest.json for base revision, document selection hashes, training parameters and export validation. Training does not establish an improvement in compassion; evaluate that separately.
ZGCM-1-7B-4bits-MLX is an open-weight model for text generation from Wu Alpha. It has 7.4B parameters and a 262,144-token context. At 16-bit it needs about 17.7 GB of GPU memory, which fits on 1x MI300X from $1.85 an hour, at the lowest prices in the SAVRN Index.
Parameters7.4B
Context262,144
Weights4.2 GB
License—
AccessOpen weights
Monthly Downloads—
Runs On
What it takes to serve ZGCM-1-7B-4bits-MLX (7.4B parameters): the memory its weights need at each precision, and the cheapest way to rent enough data-center GPUs to hold them.
| Precision | Weights | Memory needed | Cheapest setup | Per hour | Also fits |
|---|---|---|---|---|---|
| 16-bit | 14.8 GB | 17.7 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 8-bit | 7.4 GB | 8.9 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 4-bit | 3.7 GB | 4.4 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
Memory is the weights at that precision plus 20% for the runtime and a short context; a long context needs more. Prices are the lowest on-demand hourly rates in the SAVRN Index, read Oct 1, 2026.
ZGCM-1-7B-4bits-MLX on every accelerator the SAVRN Index prices, at every precision
Model Card
The publisher has not written a card for this model.
Configuration
- Architecture
- ZgcmForCausalLM
- Context length (tokens)
- 262,144
- Layers
- 32
- Hidden size
- 4,096
- Feed-forward size
- 11,008
- Attention heads
- 32
- Key/value heads
- 8
- Head dimension
- 128
- Vocabulary size
- 155,136
- Sliding window (tokens)
- 128
- RoPE base
- 1e+07
- Stored precision
- bfloat16
- Model type
- zgcm
Identity and Version
- Repository
- AlphaOxO/ZGCM-1-7B-4bits-MLX
- Publisher
- Wu Alpha
- Task
- Text generation
- Modality
- Text
- Library
- mlx
- Parameters
- 7.4B parameters
- Languages
- en
- Revision
- fba727bb96e4957ba2f2365f7a50b21829b300a2
- First published
- 2026-09-21
- Last updated
- 2026-09-21
Files and Weights
11 files, 4.2 GB in total. The weights are 1 file totalling 4.2 GB in safetensors.
Weights1 file · 4.2 GB
Configuration5 files · 85.5 KB
Tokenizer2 files · 20.2 MB
Documentation1 file · 81 B
Other1 file · 4.5 KB
Repository1 file · 1.6 KB
Every file
| File | Type | Size | SHA-256 |
|---|---|---|---|
| model.safetensors | Weights | 4.2 GB | ae5b6c6346cd |
| config.json | Configuration | 2.6 KB | — |
| configuration_zgcm.py | Configuration | 2.5 KB | — |
| generation_config.json | Configuration | 55 B | — |
| model.safetensors.index.json | Configuration | 63.2 KB | — |
| modeling_zgcm.py | Configuration | 17.2 KB | — |
| README.md | Documentation | 81 B | — |
| chat_template.jinja | Other | 4.5 KB | — |
| .gitattributes | Repository | 1.6 KB | — |
| tokenizer.json | Tokenizer | 20.2 MB | 19e773648cb4 |
| tokenizer_config.json | Tokenizer | 403 B | — |
License and Download
- License
- Not stated by the source
- Access
- Open weights, no gate
- Download size
- 4.2 GB
Download from Wu Alpha
Released by Wu Alpha through its official repository on Hugging Face.
Memory Requirements
| Precision | Weights in memory |
|---|---|
| As published | 4.2 GB |
| 16-bit | 14.8 GB |
| 8-bit | 7.4 GB |
| 4-bit | 3.7 GB |
Weights only, from the published parameter count; the key-value cache and runtime add to this.
Questions About ZGCM-1-7B-4bits-MLX
How much GPU memory does ZGCM-1-7B-4bits-MLX need?
About 17.7 GB at 16-bit and 4.4 GB at 4-bit: the weights (7.4B parameters) plus a working margin. A long context needs more.
What is the cheapest GPU to run ZGCM-1-7B-4bits-MLX on?
At 16-bit, 1x MI300X from $1.85 an hour; at 4-bit, 1x MI300X from $1.85 an hour, at the lowest on-demand prices the SAVRN Index lists.
What is ZGCM-1-7B-4bits-MLX's context length?
262,144 tokens, from the maximum position embeddings in its published configuration.
Similar Models
An experimental post-training transfer study. We applied the Vinci SFT + DPO character recipe to mistralai/Mistral-7B-Instruct-v0.3 to answer one question: does character training developed on a different model lineage transfer to this one? Apache-2.0, 7.25B, drop-in with transformers. This uses a retired base, and we are saying so first. Mistral lists Mistral 7B Instruct v0.3 as retired as of 30 March 2025 (deprecated 30 November 2024), with Ministral 3 8B as the recommended replacement. "Retired" is Mistral's own lifecycle term. The open weights remain downloadable on Hugging Face under Apache-2.0. We selected this base for continuity with our earlier experiments, not because it is…
The Mistral-7B-Instruct-v0.2 Large Language Model (LLM) is an instruct fine-tuned version of the Mistral-7B-v0.2. Mistral-7B-v0.2 has the following changes compared to Mistral-7B-v0.1 - 32k context window (vs 8k context in v0.1) - Rope-theta = 1e6 For full details of this model please read our paper and release blog post. In order to leverage instruction fine-tuning, your prompt should be surrounded by [INST] and [/INST] tokens. The very first instruction should begin with a begin of sentence id. The next instructions should not. The assistant generation will be ended by the end-of-sentence token id. This format is available as a chat template via the applychattemplate() method: - If you…
A Mistral-7B-Instruct-v0.2 checkpoint compressed with SVD-LLM to 70.0% of dense parameters, then edited by 10 of 10 rounds of iterative parameter-neutral swap selected by the gapiter rule (up to 0.1% of dense parameters per round; the full run's budget is 1.0%). This is a research artifact from a study of how SVD compression damages safety behaviour and which component-selection rule best repairs it. It is one cell of a grid over selection rules and budgets; it is not a general-purpose chat model. This checkpoint exists to measure safety/utility trade-offs under compression. Several arms in the grid are deliberately safety-degraded relative to Mistral-7B-Instruct-v0.2: compression alone…
This is the model card of a transformers model that has been pushed on the Hub. Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. Use the code below to get started with the model. Carbon emissions can be estimated using the Machine Learning Impact calculator presented in Lacoste et al. (2019).
Qwen2.5 is the latest series of Qwen large language models. For Qwen2.5, we release a number of base language models and instruction-tuned language models ranging from 0.5 to 72 billion parameters. Qwen2.5 brings the following improvements upon Qwen2: - Significantly more knowledge and has greatly improved capabilities in coding and mathematics, thanks to our specialized expert models in these domains. - Significant improvements in instruction following, generating long texts (over 8K tokens), understanding structured data (e.g, tables), and generating structured outputs especially JSON. More resilient to the diversity of system prompts, enhancing role-play implementation and…