Views
No views yet

Suggestive as Magnum models usually are, along with keeping some of the intelligence that was nice to have with the Gemma2 family.1"""<|im_start|>system
2system prompt<|im_end|>
3<|im_start|>user
4Hi there!<|im_end|>
5<|im_start|>assistant
6Nice to meet you!<|im_end|>
7<|im_start|>user
8Can I ask a question?<|im_end|>
9<|im_start|>assistant
10"""0.02 minp for the models, The model may act dumb or otherwise stupid without it.Currently, your role is {{char}}, described in detail below. As {{char}}, continue the narrative exchange with {{user}}.
<Guidelines>
• Maintain the character persona but allow it to evolve with the story.
• Be creative and proactive. Drive the story forward, introducing plotlines and events when relevant.
• All types of outputs are encouraged; respond accordingly to the narrative.
• Include dialogues, actions, and thoughts in each response.
• Utilize all five senses to describe scenarios within {{char}}'s dialogue.
• Use emotional symbols such as "!" and "~" in appropriate contexts.
• Incorporate onomatopoeia when suitable.
• Allow time for {{user}} to respond with their own input, respecting their agency.
• Act as secondary characters and NPCs as needed, and remove them when appropriate.
• When prompted for an Out of Character [OOC:] reply, answer neutrally and in plaintext, not as {{char}}.
</Guidelines>
<Forbidden>
• Using excessive literary embellishments and purple prose unless dictated by {{char}}'s persona.
• Writing for, speaking, thinking, acting, or replying as {{user}} in your response.
• Repetitive and monotonous outputs.
• Positivity bias in your replies.
• Being overly extreme or NSFW when the narrative context is inappropriate.
</Forbidden>
Follow the instructions in <Guidelines></Guidelines>, avoiding the items listed in <Forbidden></Forbidden>.
0.4.11base_model: /workspace/data/gemma-2-9b-chatml
2model_type: AutoModelForCausalLM
3tokenizer_type: AutoTokenizer
4
5plugins:
6 - axolotl.integrations.liger.LigerPlugin
7liger_rope: false
8liger_rms_norm: false
9liger_swiglu: true
10liger_cross_entropy: true
11liger_fused_linear_cross_entropy: false
12
13load_in_8bit: false
14load_in_4bit: false
15strict: false
16
17datasets:
18 - path: anthracite-org/c2_logs_16k_llama_v1.1
19 type: sharegpt
20 conversation: chatml
21 - path: NewEden/Claude-Instruct-5K
22 type: sharegpt
23 conversation: chatml
24 - path: anthracite-org/kalo-opus-instruct-22k-no-refusal
25 type: sharegpt
26 conversation: chatml
27 - path: Epiculous/SynthRP-Gens-v1.1-Filtered-n-Cleaned
28 type: sharegpt
29 conversation: chatml
30 - path: lodrick-the-lafted/kalo-opus-instruct-3k-filtered
31 type: sharegpt
32 conversation: chatml
33 - path: anthracite-org/nopm_claude_writing_fixed
34 type: sharegpt
35 conversation: chatml
36 - path: Epiculous/Synthstruct-Gens-v1.1-Filtered-n-Cleaned
37 type: sharegpt
38 conversation: chatml
39 - path: anthracite-org/kalo_opus_misc_240827
40 type: sharegpt
41 conversation: chatml
42 - path: anthracite-org/kalo_misc_part2
43 type: sharegpt
44 conversation: chatml
45chat_template: chatml
46shuffle_merged_datasets: false
47default_system_message: "You are a helpful assistant that responds to the user."
48dataset_prepared_path: /workspace/data/9b-fft-data
49val_set_size: 0.0
50output_dir: /workspace/data/9b-fft-out
51
52sequence_len: 8192
53sample_packing: true
54eval_sample_packing: false
55pad_to_sequence_len: true
56
57adapter:
58lora_model_dir:
59lora_r:
60lora_alpha:
61lora_dropout:
62lora_target_linear:
63lora_fan_in_fan_out:
64
65wandb_project: 9b-Nemo-config-fft
66wandb_entity:
67wandb_watch:
68wandb_name: attempt-01
69wandb_log_model:
70
71gradient_accumulation_steps: 4
72micro_batch_size: 1
73num_epochs: 4
74optimizer: paged_adamw_8bit
75lr_scheduler: cosine
76learning_rate: 0.00001
77
78train_on_inputs: false
79group_by_length: false
80bf16: auto
81fp16:
82tf32: false
83
84gradient_checkpointing: true
85early_stopping_patience:
86auto_resume_from_checkpoints: true
87local_rank:
88logging_steps: 1
89xformers_attention:
90flash_attention: true
91
92warmup_steps: 10
93evals_per_epoch:
94eval_table_size:
95eval_max_new_tokens:
96saves_per_epoch: 1
97debug:
98deepspeed: deepspeed_configs/zero3_bf16.json
99weight_decay: 0.001
100fsdp:
101fsdp_config:
102special_tokens:
103 pad_token: <pad>
104| Metric | Value |
|---|---|
| Avg. | 24.65 |
| IFEval (0-Shot) | 36.92 |
| BBH (3-Shot) | 34.83 |
| MATH Lvl 5 (4-Shot) | 12.54 |
| GPQA (0-shot) | 12.19 |
| MuSR (0-shot) | 17.56 |
| MMLU-PRO (5-shot) | 33.85 |