CoolFace
Modelpublic

BeaverAI/MS-2501-DPE-QwQify-v0.1-24B

sourceHugging Faceapache-2.0updated 2y agoView on Hugging Face
5likes33downloads
Model Card

BeaverAI/MS-2501-DPE-QwQify-v0.1-24B

Test model to try to give an existing model QwQ's thoughts. For this version it is ontop of `PocketDoc/Dans-PersonalityEngine-V1.2.0-24b` (a jack of all trades instruct model), which was trained ontop of `mistralai/Mistral-Small-24B-Base-2501`.

The prompt formatting and usage should be the same as with QwQ; Use ChatML, and remove the thinking from previous turns. If thoughts arent being generated automatically, add <think>\n to the start of the assistant turn.

It should follow previous model turns formatting. On first turns of the conversation you may need to regen a few times, and maybe edit the model responses for the first few turns to get it to your liking.

image/png

image/png

<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->

<img src="https://raw.githubusercontent.com/axolotl-ai-cloud/axolotl/main/image/axolotl-badge-web.png" alt="Built with Axolotl" width="200" height="32"/> <details><summary>See axolotl config</summary>

axolotl version: 0.8.0.dev0

yaml
mlflow_tracking_uri: http://127.0.0.1:7860
mlflow_experiment_name: MS-2501-DPE-QwQify-v0.1-24B-LoRA

# Hugging Face saving config
hub_model_id: BeaverAI/MS-2501-DPE-QwQify-v0.1-24B-LoRA-WS
hub_strategy: every_save

# Model checkpointing config
output_dir: ./Outputs/MS-2501-DPE-QwQify-v0.1-24B-LoRA
resume_from_checkpoint:
save_steps: 50
save_safetensors: true
save_total_limit: 3
save_only_model: false

# Model architecture config
base_model: PocketDoc/Dans-PersonalityEngine-V1.2.0-24b
model_type: MistralForCausalLM
tokenizer_type: AutoTokenizer

# Mixed precision training config
bf16: true
fp16: false
tf32: false

# Model loading config
load_in_8bit: false
load_in_4bit: false
strict: false

# Sequence config
sequence_len: 8192
min_sample_len: 256
sample_packing: true
eval_sample_packing: true
pad_to_sequence_len: true
train_on_inputs: false
group_by_length: false

# LoRA adapter config
adapter: lora
lora_model_dir:
lora_r: 128
lora_alpha: 128
lora_dropout: 0.125
peft_layers_to_transform:
peft_use_dora:
peft_use_rslora:
peft_layer_replication:
lora_target_modules:
  - gate_proj
  - down_proj
  - up_proj
  - q_proj
  - v_proj
  - k_proj
  - o_proj
lora_modules_to_save:

# Fix uninitialized tokens (such as <|start_header_id|> on the base L3 models)
fix_untrained_tokens:

# Dataset config
# https://github.com/xzuyn/axolotl/blob/came-plus-formatters/src/axolotl/prompt_strategies/customchatml-regex-last-only.py
datasets:
  - path: PJMixers-Dev/allura-org_gryphe-sonnet-3.5-charcards-names-added-qwq-all-aphrodite-Shuffled
    split: train
    type: customchatml-regex-last-only
  - path: PJMixers-Dev/anthracite-org_c2_logs_32k_llama3_qwen2_v1.3-qwq-all-aphrodite-Shuffled
    split: train
    type: customchatml-regex-last-only
  - path: PJMixers-Dev/grimulkan_aicg-logs-augmented-system-qwq-all-aphrodite-Shuffled
    split: train
    type: customchatml-regex-last-only
  - path: PJMixers-Dev/grimulkan_jannie-log-augmented-system-qwq-all-aphrodite-Shuffled
    split: train
    type: customchatml-regex-last-only
  - path: PJMixers-Dev/grimulkan_PIPPA-augmented-dedup-system-qwq-all-aphrodite-Shuffled
    split: train
    type: customchatml-regex-last-only
  - path: PJMixers-Dev/lemonilia_LimaRP-Only-NonSus-Simple-CustomShareGPT-qwq-all-aphrodite-Shuffled
    split: train
    type: customchatml-regex-last-only
  - path: PJMixers-Dev/MinervaAI_Aesir-Preview-Anon-qwq-all-aphrodite-Shuffled
    split: train
    type: customchatml-regex-last-only
  - path: PJMixers-Dev/NyxKrage_chub-logs-sharegpt-longest-CustomShareGPT-qwq-all-aphrodite-Shuffled
    split: train
    type: customchatml-regex-last-only
  - path: PJMixers-Dev/PocketDoc_Dans-Prosemaxx-Cowriter-XL-8192-shrunk-l3-qwq-all-aphrodite-Shuffled
    split: train
    type: customchatml-regex-last-only
  - path: PJMixers-Dev/PocketDoc_Dans-Personamaxx-Rainy-qwq-all-aphrodite-Shuffled
    split: train
    type: customchatml-regex-last-only
test_datasets:
  - path: PJMixers-Dev/allura-org_gryphe-sonnet-3.5-charcards-names-added-qwq-all-aphrodite-Shuffled
    split: test
    type: customchatml-regex-last-only
  - path: PJMixers-Dev/anthracite-org_c2_logs_32k_llama3_qwen2_v1.3-qwq-all-aphrodite-Shuffled
    split: test
    type: customchatml-regex-last-only
  - path: PJMixers-Dev/grimulkan_aicg-logs-augmented-system-qwq-all-aphrodite-Shuffled
    split: test
    type: customchatml-regex-last-only
  - path: PJMixers-Dev/grimulkan_jannie-log-augmented-system-qwq-all-aphrodite-Shuffled
    split: test
    type: customchatml-regex-last-only
  - path: PJMixers-Dev/grimulkan_PIPPA-augmented-dedup-system-qwq-all-aphrodite-Shuffled
    split: test
    type: customchatml-regex-last-only
  - path: PJMixers-Dev/lemonilia_LimaRP-Only-NonSus-Simple-CustomShareGPT-qwq-all-aphrodite-Shuffled
    split: test
    type: customchatml-regex-last-only
  - path: PJMixers-Dev/MinervaAI_Aesir-Preview-Anon-qwq-all-aphrodite-Shuffled
    split: test
    type: customchatml-regex-last-only
  - path: PJMixers-Dev/NyxKrage_chub-logs-sharegpt-longest-CustomShareGPT-qwq-all-aphrodite-Shuffled
    split: test
    type: customchatml-regex-last-only
  - path: PJMixers-Dev/PocketDoc_Dans-Prosemaxx-Cowriter-XL-8192-shrunk-l3-qwq-all-aphrodite-Shuffled
    split: test
    type: customchatml-regex-last-only
  - path: PJMixers-Dev/PocketDoc_Dans-Personamaxx-Rainy-qwq-all-aphrodite-Shuffled
    split: test
    type: customchatml-regex-last-only
val_set_size: 0
eval_strategy: steps
eval_steps: 50
dataset_prepared_path: ./00-Tokenized-Datasets/MS-2501-DPE-QwQify-v0.1-24B-customchatml-regex-last-only
shuffle_merged_datasets: true
dataset_processes:

# Training hyperparameters
num_epochs: 2
gradient_accumulation_steps: 1
micro_batch_size: 8  # x4 GPUs = 32
eval_batch_size: 8   # x4 GPUs = 32
warmup_steps: 0
optimizer: came_pytorch
optim_args:
optim_target_modules:
lr_scheduler: rex
learning_rate: 2e-5
cosine_min_lr_ratio:
loraplus_lr_ratio:
loraplus_lr_embedding:
weight_decay: 0.1
max_grad_norm: 1
logging_steps: 1

# Model optimization
gradient_checkpointing: unsloth
flash_attention: true
plugins:
  - axolotl.integrations.liger.LigerPlugin
  - axolotl.integrations.cut_cross_entropy.CutCrossEntropyPlugin
cut_cross_entropy: true
liger_rope: true
liger_rms_norm: true
liger_layer_norm: true
liger_glu_activation: true
liger_cross_entropy: false
liger_fused_linear_cross_entropy: false
lora_mlp_kernel: false
lora_qkv_kernel: false
lora_o_kernel: false

# DeepSpeed
deepspeed: deepspeed_configs/zero3_bf16.json

# Garbage Collection
gc_steps: 1

# Debug config
debug: true
seed: 42

# Token config
special_tokens:
  bos_token: "<s>"
  eos_token: "<|im_end|>"
  pad_token: "<pad>"
tokens:

</details><br>

MS-2501-DPE-QwQify-v0.1-24B-LoRA-WS

This model is a fine-tuned version of PocketDoc/Dans-PersonalityEngine-V1.2.0-24b on the PJMixers-Dev/allura-orggryphe-sonnet-3.5-charcards-names-added-qwq-all-aphrodite-Shuffled, the PJMixers-Dev/anthracite-orgc2logs32kllama3qwen2v1.3-qwq-all-aphrodite-Shuffled, the PJMixers-Dev/grimulkanaicg-logs-augmented-system-qwq-all-aphrodite-Shuffled, the PJMixers-Dev/grimulkanjannie-log-augmented-system-qwq-all-aphrodite-Shuffled, the PJMixers-Dev/grimulkanPIPPA-augmented-dedup-system-qwq-all-aphrodite-Shuffled, the PJMixers-Dev/lemoniliaLimaRP-Only-NonSus-Simple-CustomShareGPT-qwq-all-aphrodite-Shuffled, the PJMixers-Dev/MinervaAIAesir-Preview-Anon-qwq-all-aphrodite-Shuffled, the PJMixers-Dev/NyxKragechub-logs-sharegpt-longest-CustomShareGPT-qwq-all-aphrodite-Shuffled, the PJMixers-Dev/PocketDocDans-Prosemaxx-Cowriter-XL-8192-shrunk-l3-qwq-all-aphrodite-Shuffled and the PJMixers-Dev/PocketDoc_Dans-Personamaxx-Rainy-qwq-all-aphrodite-Shuffled datasets. It achieves the following results on the evaluation set:

  • —Loss: 1.1949

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

  • —learning_rate: 2e-05
  • —trainbatchsize: 8
  • —evalbatchsize: 8
  • —seed: 42
  • —distributed_type: multi-GPU
  • —num_devices: 4
  • —totaltrainbatch_size: 32
  • —totalevalbatch_size: 32
  • —optimizer: Use OptimizerNames.ADAMWHF with betas=(0.9,0.999) and epsilon=1e-08 and optimizerargs=No additional optimizer arguments
  • —lrschedulertype: cosine
  • —num_epochs: 2.0

Training results

Training LossEpochStepValidation Loss
1.99250.001911.9225
1.42280.0936501.4329
1.34730.18731001.3722
1.32590.28091501.3414
1.27950.37452001.3199
1.28170.46822501.3029
1.23650.56183001.2910
1.21340.65543501.2803
1.26550.74914001.2700
1.22970.84274501.2614
1.1780.93635001.2524
1.15251.03005501.2467
1.17511.12366001.2411
1.2161.21726501.2366
1.17061.31097001.2302
1.13631.40457501.2256
1.15631.49818001.2194
1.15591.59188501.2147
1.12631.68549001.2090
1.0991.77909501.2038
1.17861.872710001.1994
1.10571.966310501.1949

Framework versions

  • —PEFT 0.14.0
  • —Transformers 4.49.0
  • —Pytorch 2.6.0+cu124
  • —Datasets 3.2.0
  • —Tokenizers 0.21.1