Skip to content

Commit 57b302b

Browse files
hanbitmythshanbitmyths
authored
Add ministral-3-3b-2512 model into Foundry Local (#5089)
* Add ministral-3-3b-instruct-2512 into Foundry Local * Update description
1 parent a7e2f7c commit 57b302b

16 files changed

Lines changed: 243 additions & 0 deletions

File tree

‎assets/models/foundrylocal/ministral-3-3b-instruct-2512-cuda-gpu/asset.yaml‎

Lines changed: 4 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,4 @@
1+
extra_config: model.yaml
2+
spec: spec.yaml
3+
type: model
4+
categories: ["Local"]

‎assets/models/foundrylocal/ministral-3-3b-instruct-2512-cuda-gpu/description.md‎

Lines changed: 15 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,15 @@
1+
This model is an optimized version of Ministral-3-3B-Instruct-2512 for local inference on CUDA GPUs. This optimized model is published here in ONNX format to run on CUDA-capable GPU devices, with the precision best suited to this target.
2+
3+
# ONNX Models
4+
Here are some of the optimized configurations we have added:
5+
1. ONNX model for CUDA GPU using RTN quantization.
6+
7+
# Model Description
8+
- **Developed by:** Microsoft
9+
- **Model type:** ONNX
10+
- **License:** apache-2.0
11+
- **Model Description:** This is a conversion of Ministral-3-3B-Instruct-2512 for local inference.
12+
- **Disclaimer:** Model is only an optimization of the base model, any risk associated with the model is the responsibility of the user of the model. Please verify and test for your scenarios. There may be a slight difference in output from the base model with the optimizations applied. Note that optimizations applied are distinct from fine tuning and thus do not alter the intended uses or capabilities of the model.
13+
14+
# Base Model Information
15+
See Hugging Face model [Ministral-3-3B-Instruct-2512](https://huggingface.co/mistralai/Ministral-3-3B-Instruct-2512) for details.

‎assets/models/foundrylocal/ministral-3-3b-instruct-2512-cuda-gpu/model.yaml‎

Lines changed: 8 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,8 @@
1+
path:
2+
container_name: models
3+
container_path: foundrylocal/models/ministral-3-3b-Instruct-2512/onnx/cuda/v1
4+
storage_name: foundrylocalassetdata
5+
type: azureblob
6+
publish:
7+
description: description.md
8+
type: custom_model

‎assets/models/foundrylocal/ministral-3-3b-instruct-2512-cuda-gpu/spec.yaml‎

Lines changed: 42 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,42 @@
1+
$schema: https://azuremlschemas.azureedge.net/latest/model.schema.json
2+
name: ministral-3-3b-instruct-2512-cuda-gpu
3+
version: 1
4+
isArchived: true
5+
path: ./
6+
tags:
7+
foundryLocal: ""
8+
license: "apache-2.0"
9+
licenseDescription: "This model is provided under the License Terms available at <https://www.apache.org/licenses/LICENSE-2.0.html>."
10+
author: Microsoft
11+
inputModalities: "text,image"
12+
outputModalities: "text"
13+
task: vision-language-chat
14+
maxOutputTokens: 2048
15+
alias: ministral-3-3b-instruct-2512
16+
directoryPath: v1
17+
promptTemplate: "{\"system\": \"[SYSTEM_PROMPT]{Content}[/SYSTEM_PROMPT]\", \"user\": \"[INST]{Content}[/INST]\", \"assistant\": \"{Content}</s>\", \"prompt\": \"[INST]{Content}[/INST]\"}"
18+
supportsToolCalling: "true"
19+
toolCallStart: "[TOOL_CALLS]"
20+
toolCallEnd: "</s>"
21+
toolRegisterStart: "[AVAILABLE_TOOLS]"
22+
toolRegisterEnd: "[/AVAILABLE_TOOLS]"
23+
toolResponseStart: "[TOOL_RESULTS]"
24+
toolResponseEnd: "[/TOOL_RESULTS]"
25+
capabilities: "tool-calling"
26+
supportsReasoning: ""
27+
reasoningStart: ""
28+
reasoningEnd: ""
29+
contextLength: 262144
30+
minFLVersion: "1.2.0"
31+
disable-maap: "true"
32+
type: custom_model
33+
variantInfo:
34+
parents:
35+
- assetId: azureml://registries/azureml/models/ministral-3-3b-instruct-2512/versions/1
36+
variantMetadata:
37+
modelType: 'ONNX'
38+
quantization: ['RTN']
39+
device: 'gpu'
40+
executionProvider: 'CUDAExecutionProvider'
41+
fileSizeBytes: 3857772715
42+
vRamFootprintBytes: 3857772715

‎assets/models/foundrylocal/ministral-3-3b-instruct-2512-generic-cpu/asset.yaml‎

Lines changed: 4 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,4 @@
1+
extra_config: model.yaml
2+
spec: spec.yaml
3+
type: model
4+
categories: ["Local"]

‎assets/models/foundrylocal/ministral-3-3b-instruct-2512-generic-cpu/description.md‎

Lines changed: 15 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,15 @@
1+
This model is an optimized version of Ministral-3-3B-Instruct-2512 for local inference on CPU and mobile devices. This optimized model is published here in ONNX format to run on CPU and mobile targets, with the precision best suited to this target.
2+
3+
# ONNX Models
4+
Here are some of the optimized configurations we have added:
5+
1. ONNX model for CPU and mobile using RTN quantization.
6+
7+
# Model Description
8+
- **Developed by:** Microsoft
9+
- **Model type:** ONNX
10+
- **License:** apache-2.0
11+
- **Model Description:** This is a conversion of Ministral-3-3B-Instruct-2512 for local inference.
12+
- **Disclaimer:** Model is only an optimization of the base model, any risk associated with the model is the responsibility of the user of the model. Please verify and test for your scenarios. There may be a slight difference in output from the base model with the optimizations applied. Note that optimizations applied are distinct from fine tuning and thus do not alter the intended uses or capabilities of the model.
13+
14+
# Base Model Information
15+
See Hugging Face model [Ministral-3-3B-Instruct-2512](https://huggingface.co/mistralai/Ministral-3-3B-Instruct-2512) for details.

‎assets/models/foundrylocal/ministral-3-3b-instruct-2512-generic-cpu/model.yaml‎

Lines changed: 8 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,8 @@
1+
path:
2+
container_name: models
3+
container_path: foundrylocal/models/ministral-3-3b-Instruct-2512/onnx/cpu_and_mobile/v1
4+
storage_name: foundrylocalassetdata
5+
type: azureblob
6+
publish:
7+
description: description.md
8+
type: custom_model

‎assets/models/foundrylocal/ministral-3-3b-instruct-2512-generic-cpu/spec.yaml‎

Lines changed: 42 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,42 @@
1+
$schema: https://azuremlschemas.azureedge.net/latest/model.schema.json
2+
name: ministral-3-3b-instruct-2512-generic-cpu
3+
version: 1
4+
isArchived: true
5+
path: ./
6+
tags:
7+
foundryLocal: ""
8+
license: "apache-2.0"
9+
licenseDescription: "This model is provided under the License Terms available at <https://www.apache.org/licenses/LICENSE-2.0.html>."
10+
author: Microsoft
11+
inputModalities: "text,image"
12+
outputModalities: "text"
13+
task: vision-language-chat
14+
maxOutputTokens: 2048
15+
alias: ministral-3-3b-instruct-2512
16+
directoryPath: v1
17+
promptTemplate: "{\"system\": \"[SYSTEM_PROMPT]{Content}[/SYSTEM_PROMPT]\", \"user\": \"[INST]{Content}[/INST]\", \"assistant\": \"{Content}</s>\", \"prompt\": \"[INST]{Content}[/INST]\"}"
18+
supportsToolCalling: "true"
19+
toolCallStart: "[TOOL_CALLS]"
20+
toolCallEnd: "</s>"
21+
toolRegisterStart: "[AVAILABLE_TOOLS]"
22+
toolRegisterEnd: "[/AVAILABLE_TOOLS]"
23+
toolResponseStart: "[TOOL_RESULTS]"
24+
toolResponseEnd: "[/TOOL_RESULTS]"
25+
capabilities: "tool-calling"
26+
supportsReasoning: ""
27+
reasoningStart: ""
28+
reasoningEnd: ""
29+
contextLength: 262144
30+
minFLVersion: "1.2.0"
31+
disable-maap: "true"
32+
type: custom_model
33+
variantInfo:
34+
parents:
35+
- assetId: azureml://registries/azureml/models/ministral-3-3b-instruct-2512/versions/1
36+
variantMetadata:
37+
modelType: 'ONNX'
38+
quantization: ['RTN']
39+
device: 'cpu'
40+
executionProvider: 'CPUExecutionProvider'
41+
fileSizeBytes: 4919876920
42+
vRamFootprintBytes: 4919876920

‎assets/models/foundrylocal/ministral-3-3b-instruct-2512-generic-gpu/asset.yaml‎

Lines changed: 4 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,4 @@
1+
extra_config: model.yaml
2+
spec: spec.yaml
3+
type: model
4+
categories: ["Local"]

‎assets/models/foundrylocal/ministral-3-3b-instruct-2512-generic-gpu/description.md‎

Lines changed: 15 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,15 @@
1+
This model is an optimized version of Ministral-3-3B-Instruct-2512 for local inference on WebGPU-capable devices. This optimized model is published here in ONNX format to run with the WebGPU execution provider, with the precision best suited to this target.
2+
3+
# ONNX Models
4+
Here are some of the optimized configurations we have added:
5+
1. ONNX model for WebGPU using RTN quantization.
6+
7+
# Model Description
8+
- **Developed by:** Microsoft
9+
- **Model type:** ONNX
10+
- **License:** apache-2.0
11+
- **Model Description:** This is a conversion of Ministral-3-3B-Instruct-2512 for local inference.
12+
- **Disclaimer:** Model is only an optimization of the base model, any risk associated with the model is the responsibility of the user of the model. Please verify and test for your scenarios. There may be a slight difference in output from the base model with the optimizations applied. Note that optimizations applied are distinct from fine tuning and thus do not alter the intended uses or capabilities of the model.
13+
14+
# Base Model Information
15+
See Hugging Face model [Ministral-3-3B-Instruct-2512](https://huggingface.co/mistralai/Ministral-3-3B-Instruct-2512) for details.

0 commit comments

Comments
 (0)