Hugging Face's logo

TareksLab
/
EXPERIMENTAL-UPSCALE-105B

Text Generation
Transformers
Safetensors
llama
mergekit
Merge
conversational
text-generation-inference
Model card Files Community

You need to agree to share your contact information to access this model

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

Log in or Sign Up to review the conditions and access this model content.

merge

This is a merge of pre-trained language models created using mergekit.

Merge Details

Merge Method

This model was merged using the Passthrough merge method.

Models Merged

The following models were included in the merge:

Configuration

The following YAML configuration was used to produce this model:

slices:
- sources:
  - layer_range: [0, 4]
    model: Tarek07/Legion-V2.1-LLaMa-70B
- sources:
  - layer_range: [2, 4]
    model: Tarek07/Legion-V2.1-LLaMa-70B
    parameters:
      scale:
      - filter: o_proj
        value: 0
      - filter: down_proj
        value: 0
      - value: 1
- sources:
  - layer_range: [4, 8]
    model: Tarek07/Legion-V2.1-LLaMa-70B
- sources:
  - layer_range: [6, 8]
    model: Tarek07/Legion-V2.1-LLaMa-70B
    parameters:
      scale:
      - filter: o_proj
        value: 0
      - filter: down_proj
        value: 0
      - value: 1
- sources:
  - layer_range: [8, 12]
    model: Tarek07/Legion-V2.1-LLaMa-70B
- sources:
  - layer_range: [10, 12]
    model: Tarek07/Legion-V2.1-LLaMa-70B
    parameters:
      scale:
      - filter: o_proj
        value: 0
      - filter: down_proj
        value: 0
      - value: 1
- sources:
  - layer_range: [12, 16]
    model: Tarek07/Legion-V2.1-LLaMa-70B
- sources:
  - layer_range: [14, 16]
    model: Tarek07/Legion-V2.1-LLaMa-70B
    parameters:
      scale:
      - filter: o_proj
        value: 0
      - filter: down_proj
        value: 0
      - value: 1
- sources:
  - layer_range: [16, 20]
    model: Tarek07/Legion-V2.1-LLaMa-70B
- sources:
  - layer_range: [18, 20]
    model: Tarek07/Legion-V2.1-LLaMa-70B
    parameters:
      scale:
      - filter: o_proj
        value: 0
      - filter: down_proj
        value: 0
      - value: 1
- sources:
  - layer_range: [20, 24]
    model: Tarek07/Legion-V2.1-LLaMa-70B
- sources:
  - layer_range: [22, 24]
    model: Tarek07/Legion-V2.1-LLaMa-70B
    parameters:
      scale:
      - filter: o_proj
        value: 0
      - filter: down_proj
        value: 0
      - value: 1
- sources:
  - layer_range: [24, 28]
    model: Tarek07/Legion-V2.1-LLaMa-70B
- sources:
  - layer_range: [26, 28]
    model: Tarek07/Legion-V2.1-LLaMa-70B
    parameters:
      scale:
      - filter: o_proj
        value: 0
      - filter: down_proj
        value: 0
      - value: 1
- sources:
  - layer_range: [28, 32]
    model: Tarek07/Legion-V2.1-LLaMa-70B
- sources:
  - layer_range: [30, 32]
    model: Tarek07/Legion-V2.1-LLaMa-70B
    parameters:
      scale:
      - filter: o_proj
        value: 0
      - filter: down_proj
        value: 0
      - value: 1
- sources:
  - layer_range: [32, 36]
    model: Tarek07/Legion-V2.1-LLaMa-70B
- sources:
  - layer_range: [34, 36]
    model: Tarek07/Legion-V2.1-LLaMa-70B
    parameters:
      scale:
      - filter: o_proj
        value: 0
      - filter: down_proj
        value: 0
      - value: 1
- sources:
  - layer_range: [36, 40]
    model: Tarek07/Legion-V2.1-LLaMa-70B
- sources:
  - layer_range: [38, 40]
    model: Tarek07/Legion-V2.1-LLaMa-70B
    parameters:
      scale:
      - filter: o_proj
        value: 0
      - filter: down_proj
        value: 0
      - value: 1
- sources:
  - layer_range: [40, 44]
    model: Tarek07/Legion-V2.1-LLaMa-70B
- sources:
  - layer_range: [42, 44]
    model: Tarek07/Legion-V2.1-LLaMa-70B
    parameters:
      scale:
      - filter: o_proj
        value: 0
      - filter: down_proj
        value: 0
      - value: 1
- sources:
  - layer_range: [44, 48]
    model: Tarek07/Legion-V2.1-LLaMa-70B
- sources:
  - layer_range: [46, 48]
    model: Tarek07/Legion-V2.1-LLaMa-70B
    parameters:
      scale:
      - filter: o_proj
        value: 0
      - filter: down_proj
        value: 0
      - value: 1
- sources:
  - layer_range: [48, 52]
    model: Tarek07/Legion-V2.1-LLaMa-70B
- sources:
  - layer_range: [50, 52]
    model: Tarek07/Legion-V2.1-LLaMa-70B
    parameters:
      scale:
      - filter: o_proj
        value: 0
      - filter: down_proj
        value: 0
      - value: 1
- sources:
  - layer_range: [52, 56]
    model: Tarek07/Legion-V2.1-LLaMa-70B
- sources:
  - layer_range: [54, 56]
    model: Tarek07/Legion-V2.1-LLaMa-70B
    parameters:
      scale:
      - filter: o_proj
        value: 0
      - filter: down_proj
        value: 0
      - value: 1
- sources:
  - layer_range: [56, 60]
    model: Tarek07/Legion-V2.1-LLaMa-70B
- sources:
  - layer_range: [58, 60]
    model: Tarek07/Legion-V2.1-LLaMa-70B
    parameters:
      scale:
      - filter: o_proj
        value: 0
      - filter: down_proj
        value: 0
      - value: 1
- sources:
  - layer_range: [60, 64]
    model: Tarek07/Legion-V2.1-LLaMa-70B
- sources:
  - layer_range: [62, 64]
    model: Tarek07/Legion-V2.1-LLaMa-70B
    parameters:
      scale:
      - filter: o_proj
        value: 0
      - filter: down_proj
        value: 0
      - value: 1
- sources:
  - layer_range: [64, 68]
    model: Tarek07/Legion-V2.1-LLaMa-70B
- sources:
  - layer_range: [66, 68]
    model: Tarek07/Legion-V2.1-LLaMa-70B
    parameters:
      scale:
      - filter: o_proj
        value: 0
      - filter: down_proj
        value: 0
      - value: 1
- sources:
  - layer_range: [68, 72]
    model: Tarek07/Legion-V2.1-LLaMa-70B
- sources:
  - layer_range: [70, 72]
    model: Tarek07/Legion-V2.1-LLaMa-70B
    parameters:
      scale:
      - filter: o_proj
        value: 0
      - filter: down_proj
        value: 0
      - value: 1
- sources:
  - layer_range: [72, 76]
    model: Tarek07/Legion-V2.1-LLaMa-70B
- sources:
  - layer_range: [74, 76]
    model: Tarek07/Legion-V2.1-LLaMa-70B
    parameters:
      scale:
      - filter: o_proj
        value: 0
      - filter: down_proj
        value: 0
      - value: 1
- sources:
  - layer_range: [76, 80]
    model: Tarek07/Legion-V2.1-LLaMa-70B
- sources:
  - layer_range: [78, 80]
    model: Tarek07/Legion-V2.1-LLaMa-70B
    parameters:
      scale:
      - filter: o_proj
        value: 0
      - filter: down_proj
        value: 0
      - value: 1
merge_method: passthrough
dtype: bfloat16
Downloads last month
0
Safetensors
Model size
105B params
Tensor type
BF16
·
Inference Providers NEW
Text Generation
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for TareksLab/EXPERIMENTAL-UPSCALE-105B

Base model

Tarek07/Legion-V2.1-LLaMa-70B
Finetuned
(1)
this model