mixtape_core/models/
llama.rs

1//! Meta Llama models
2
3use super::define_model;
4
5// =============================================================================
6// Llama 4 Models
7// =============================================================================
8
9define_model!(
10    /// Llama 4 Scout 17B - Efficient MoE model with 10M context
11    Llama4Scout17B {
12        display_name: "Llama 4 Scout 17B",
13        bedrock_id: "meta.llama4-scout-17b-instruct-v1:0",
14        context_tokens: 10_000_000,
15        output_tokens: 4_096
16    }
17);
18
19define_model!(
20    /// Llama 4 Maverick 17B - Larger MoE model with 1M context
21    Llama4Maverick17B {
22        display_name: "Llama 4 Maverick 17B",
23        bedrock_id: "meta.llama4-maverick-17b-instruct-v1:0",
24        context_tokens: 1_000_000,
25        output_tokens: 4_096
26    }
27);
28
29// =============================================================================
30// Llama 3.3 Models
31// =============================================================================
32
33define_model!(
34    /// Llama 3.3 70B Instruct - Latest Llama 3.x flagship
35    Llama3_3_70B {
36        display_name: "Llama 3.3 70B",
37        bedrock_id: "meta.llama3-3-70b-instruct-v1:0",
38        context_tokens: 128_000,
39        output_tokens: 4_096
40    }
41);
42
43// =============================================================================
44// Llama 3.2 Models
45// =============================================================================
46
47define_model!(
48    /// Llama 3.2 90B Instruct - Large multimodal model
49    Llama3_2_90B {
50        display_name: "Llama 3.2 90B",
51        bedrock_id: "meta.llama3-2-90b-instruct-v1:0",
52        context_tokens: 128_000,
53        output_tokens: 4_096
54    }
55);
56
57define_model!(
58    /// Llama 3.2 11B Instruct - Medium multimodal model
59    Llama3_2_11B {
60        display_name: "Llama 3.2 11B",
61        bedrock_id: "meta.llama3-2-11b-instruct-v1:0",
62        context_tokens: 128_000,
63        output_tokens: 4_096
64    }
65);
66
67define_model!(
68    /// Llama 3.2 3B Instruct - Efficient small model
69    Llama3_2_3B {
70        display_name: "Llama 3.2 3B",
71        bedrock_id: "meta.llama3-2-3b-instruct-v1:0",
72        context_tokens: 128_000,
73        output_tokens: 4_096
74    }
75);
76
77define_model!(
78    /// Llama 3.2 1B Instruct - Lightweight model for edge deployment
79    Llama3_2_1B {
80        display_name: "Llama 3.2 1B",
81        bedrock_id: "meta.llama3-2-1b-instruct-v1:0",
82        context_tokens: 128_000,
83        output_tokens: 4_096
84    }
85);
86
87// =============================================================================
88// Llama 3.1 Models
89// =============================================================================
90
91define_model!(
92    /// Llama 3.1 405B Instruct - Largest open-weights model
93    Llama3_1_405B {
94        display_name: "Llama 3.1 405B",
95        bedrock_id: "meta.llama3-1-405b-instruct-v1:0",
96        context_tokens: 128_000,
97        output_tokens: 4_096
98    }
99);
100
101define_model!(
102    /// Llama 3.1 70B Instruct - High capability model
103    Llama3_1_70B {
104        display_name: "Llama 3.1 70B",
105        bedrock_id: "meta.llama3-1-70b-instruct-v1:0",
106        context_tokens: 128_000,
107        output_tokens: 4_096
108    }
109);
110
111define_model!(
112    /// Llama 3.1 8B Instruct - Efficient general purpose model
113    Llama3_1_8B {
114        display_name: "Llama 3.1 8B",
115        bedrock_id: "meta.llama3-1-8b-instruct-v1:0",
116        context_tokens: 128_000,
117        output_tokens: 4_096
118    }
119);
mixtape_core/models/llama.rs

mixtape_core/models/
llama.rs
