yammer 0.7.0

yammer provides an ollama-compatible client library.
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182

use reqwest::RequestBuilder;

////////////////////////////////////////// GenerateRequest /////////////////////////////////////////

/// Generate a response to a prompt.
#[derive(Clone, Debug, Eq, PartialEq, serde::Deserialize, serde::Serialize)]
pub struct GenerateRequest {
    /// The name of the ollama model to use from the ollama library.
    pub model: String,

    /// The prompt to provide to the model.  This is the text that the model will use to generate a
    /// response.
    pub prompt: String,

    /// The suffix to append to the prompt.  This is useful for generating a response that is a
    /// continuation of the prompt.
    pub suffix: String,

    /// A list of base64-encoded images to supply to the model.
    pub images: Option<Vec<String>>,

    /// The format to return the response in.  If provided, this must be "json".
    #[serde(skip_serializing_if = "Option::is_none")]
    pub format: Option<String>,

    /// The system to use for the response.
    #[serde(skip_serializing_if = "Option::is_none")]
    pub system: Option<String>,

    /// The template to use for the response.
    #[serde(skip_serializing_if = "Option::is_none")]
    pub template: Option<String>,

    /// Should this response stream?
    #[serde(skip_serializing_if = "Option::is_none")]
    pub stream: Option<bool>,

    /// Should this response be raw?
    #[serde(skip_serializing_if = "Option::is_none")]
    pub raw: Option<bool>,

    /// How long to hold the model in memory for once the request completes.
    #[serde(skip_serializing_if = "Option::is_none")]
    pub keep_alive: Option<String>,

    /// Additional options to pass to the model.
    pub options: serde_json::Value,
}

impl Default for GenerateRequest {
    fn default() -> Self {
        Self {
            model: "gemma2".to_string(),
            prompt: "42".to_string(),
            suffix: "".to_string(),
            images: None,
            format: None,
            system: None,
            template: None,
            stream: None,
            raw: None,
            keep_alive: None,
            options: serde_json::json!({ "num_ctx": 12288 }),
        }
    }
}

impl GenerateRequest {
    /// Create a new RequestBuilder for this generate request.
    pub fn make_request(&self, ollama_host: &str) -> RequestBuilder {
        reqwest::Client::new()
            .post(format!("{}/api/generate", ollama_host))
            .json(self)
    }
}

///////////////////////////////////////// GenerateResponse /////////////////////////////////////////

/// A response to a generate request.
#[derive(Clone, Debug, serde::Deserialize, serde::Serialize)]
pub struct GenerateResponse {
    /// The name of the model used to generate the response.
    pub model: String,
    /// The time the response was created.
    pub created_at: String,
    /// The response generated by the model.
    pub response: String,
    /// Whether the response is done.
    pub done: bool,
    /// The duration of the response.
    pub total_duration: Option<f64>,
    /// The duration of loading the model.
    pub load_duration: Option<f64>,
    /// The number of tokens counted in the prompt.
    pub prompt_eval_count: Option<f64>,
    /// The duration of the prompt evaluation.
    pub prompt_eval_duration: Option<f64>,
    /// The number of tokens counted in the response.
    pub eval_count: Option<f64>,
    /// The duration of the response evaluation.
    pub eval_duration: Option<f64>,
    /// The context for a future generate call.
    pub context: Vec<f64>,
}

//////////////////////////////////////////// ChatMessage ///////////////////////////////////////////

/// A message sent or received in a chat.
#[derive(Clone, Debug, Eq, PartialEq, serde::Deserialize, serde::Serialize)]
pub struct ChatMessage {
    /// The role of the message sender.
    pub role: String,
    /// The content of the message.
    pub content: String,
    /// The images attached to the message.
    #[serde(skip_serializing_if = "Option::is_none")]
    pub images: Option<Vec<String>>,
    /// The tool calls made by the message.
    #[serde(skip_serializing_if = "Option::is_none")]
    pub tool_calls: Option<Vec<serde_json::Value>>,
}

//////////////////////////////////////////// ChatRequest ///////////////////////////////////////////

/// A request to chat with a model.
#[derive(Clone, Debug, serde::Deserialize, serde::Serialize)]
pub struct ChatRequest {
    /// The name of the ollama model to use from the ollama library.
    pub model: String,
    /// The chat messages to send to the model.
    pub messages: Vec<ChatMessage>,
    /// The tools available to use in the chat.
    #[serde(skip_serializing_if = "Option::is_none")]
    pub tools: Option<serde_json::Value>,
    /// The format to return the response in.  If provided, this must be "json".
    #[serde(skip_serializing_if = "Option::is_none")]
    pub format: Option<String>,
    /// Should this response stream?
    #[serde(skip_serializing_if = "Option::is_none")]
    pub stream: Option<bool>,
    /// How long to keep the model alive for after the request completes.
    #[serde(skip_serializing_if = "Option::is_none")]
    pub keep_alive: Option<String>,
    /// Additional options to pass to the model.
    pub options: serde_json::Value,
}

impl ChatRequest {
    /// Create a new RequestBuilder for this chat request.
    pub fn make_request(&self, ollama_host: &str) -> RequestBuilder {
        reqwest::Client::new()
            .post(format!("{}/api/chat", ollama_host))
            .json(self)
    }
}

/////////////////////////////////////////// ChatResponse ///////////////////////////////////////////

/// A response to a chat request.
#[derive(Clone, Debug, serde::Deserialize, serde::Serialize)]
pub struct ChatResponse {
    /// The name of the model used to generate the response.
    pub model: String,
    /// When the response was created.
    pub created_at: String,
    /// The messages generated by the model.
    pub message: ChatMessage,
    /// Whether the response is done.
    pub done: bool,
    /// The duration of the response.
    pub total_duration: Option<f64>,
    /// The duration of loading the model.
    pub load_duration: Option<f64>,
    /// The number of tokens counted in the prompt.
    pub prompt_eval_count: Option<f64>,
    /// The duration of the prompt evaluation.
    pub prompt_eval_duration: Option<f64>,
    /// The number of tokens counted in the response.
    pub eval_count: Option<f64>,
    /// The duration of the response evaluation.
    pub eval_duration: Option<f64>,
}