output_merge_multiSFT_v6_1
This is a merge of pre-trained language models created using mergekit.
Merge Details
Merge Method
This model was merged using the SLERP merge method.
Models Merged
The following models were included in the merge:
- /workspace/storage-shared/cnm/embedding/code/llama_1B_legal_retrieval_v2_2.3M_clean_150726_finetune2_07150726/checkpoint-8000
- /workspace/storage-shared/cnm/huyhq21_v2/mergekit/output_slerp_1B_data_v2_clean_v6
Configuration
The following YAML configuration was used to produce this model:
# slices:
# - sources:
# - model: psmathur/orca_mini_v3_13b
# layer_range: [0, 40]
# - model: garage-bAInd/Platypus2-13B
# layer_range: [0, 40]
# # or, the equivalent models: syntax:
# # models:
# # - model: psmathur/orca_mini_v3_13b
# # - model: garage-bAInd/Platypus2-13B
# merge_method: slerp
# base_model: psmathur/orca_mini_v3_13b
# parameters:
# t:
# - filter: self_attn
# value: [0, 0.5, 0.3, 0.7, 1]
# - filter: mlp
# value: [1, 0.5, 0.7, 0.3, 0]
# - value: 0.5 # fallback for rest of tensors
# dtype: float16
# slices:
models:
- model: /workspace/storage-shared/cnm/embedding/code/llama_1B_legal_retrieval_v2_2.3M_clean_150726_finetune2_07150726/checkpoint-8000 # Mô hình A (Base gốc)
# layer_range: [0, 36]
- model: /workspace/storage-shared/cnm/huyhq21_v2/mergekit/output_slerp_1B_data_v2_clean_v6 # Mô hình B (Đã Finetune)
# layer_range: [0, 36]
merge_method: slerp
base_model: /workspace/storage-shared/cnm/huyhq21_v2/mergekit/output_slerp_1B_data_v2_clean_v6 # Mô hình B (Đã Finetune)
parameters:
t:
#- filter: self_attn
- filter: layers
value: [0.2, 0.3, 0.5, 0.6, 0.7, 0.8]
#- filter: mlp
# value: [0.2, 0.3, 0.5, 0.6, 0.7, 0.8]
- value: 0.5 # fallback cho layernorm
dtype: bfloat16
# models:
# - model: /workspace/storage-shared/cnm/cuongnq23/project_embeding/models/Qwen/Qwen3-Embedding-8B
# - model: /workspace/storage-shared/cnm/embedding/code/qwen8b_SFT_kalm_vnlaw_3M/checkpoint-20000
# merge_method: slerp
# base_model: /workspace/storage-shared/cnm/cuongnq23/project_embeding/models/Qwen/Qwen3-Embedding-8B
# parameters:
# t:
# # Trượt dốc từ 10% (đáy) lên 90% (đỉnh)
# - value: [0.1, 0.3, 0.5, 0.7, 0.9]
# dtype: bfloat16