nateraw/wizard-mega-13b-awq 🔢📝 → 📝

▶️ 5.5K runs 📅 Sep 2023 ⚙️ Cog 0.8.6 🔗 GitHub
code-generation question-answering text-generation

About

wizard-mega-13b quantized with AWQ and served with vLLM

Example Output

Output

Great idea! Here's a short itinerary for your dog's birthday party:

  1. Introduction: Introduce your dog to the guests and explain the plan for the day.
  2. Games and Activities: Set up some fun games and activities for the dogs to play, such as a scavenger hunt, obstacle course, or agility training.
  3. Cake and Presents: Cut the dog-friendly cake and let each dog have a piece. Also, set out presents for the dogs to open and play with.
  4. Photo Opportunity: Take a group photo with all the dogs and guests.
  5. Treats and Refreshments: Provide water and treats for the dogs and guests to enjoy.
  6. Relaxation: End the party with a relaxing activity, such as a massage or belly rub for the dogs.
  7. Goodbye: Thank everyone for coming and send them home with some dog-themed party favors.

I hope this itinerary helps you plan a memorable birthday party for your dog!

Performance Metrics

5.95s Prediction Time
79.51s Total Time
All Input Parameters 
{
  "top_k": 50,
  "top_p": 0.95,
  "message": "write me a short itinerary for my dog's birthday party.",
  "temperature": 0.8,
  "max_new_tokens": 512,
  "presence_penalty": 1
}
Input Parameters
top_k Type: integerDefault: 50
The number of highest probability tokens to consider for generating the output. If > 0, only keep the top k tokens with highest probability (top-k filtering).
top_p Type: numberDefault: 1
A probability threshold for generating the output. If < 1.0, only keep the top tokens with cumulative probability >= top_p (nucleus filtering). Nucleus filtering is described in Holtzman et al. (http://arxiv.org/abs/1904.09751).
message (required) Type: string
temperature Type: numberDefault: 1
The value used to modulate the next token probabilities.
max_new_tokens Type: integerDefault: 512
The maximum number of tokens the model should generate as output.
presence_penalty Type: numberDefault: 1.15
Presence penalty
Output Schema

Output

Type: string

Example Execution Logs 
Processed prompts:   0%|          | 0/2 [00:00<?, ?it/s]
Processed prompts:  50%|█████     | 1/2 [00:04<00:04,  4.17s/it]
Processed prompts: 100%|██████████| 2/2 [00:05<00:00,  2.69s/it]
Processed prompts: 100%|██████████| 2/2 [00:05<00:00,  2.91s/it]
Generated 241 tokens in 5.8255774974823 seconds.
Version Details
Version ID
a4be2a7c75e51c53b22167d44de3333436f1aa9253a201d2619cf74286478599
Version Created
September 26, 2023
Run on Replicate →